风控模型开发验证报告

实验 ID: DEV-mvyzwkut-49fe · 生成时间: 2026-11-02 16:27 (GMT+8) · 报告结构:标准14节模型开发验证模板

00 模型概要信息

项目内容
模型名称exp_DEV-mvyzwkut-49fe
开发完成日期2026-11-02
模型开发人员(人工填写)
模型开发人员邮件地址(人工填写)
模型验证人员(人工填写)
模型验证人员邮件地址(人工填写)
模型应用环节inloan_warning
模型应用频率(人工填写)
模型开发背景及业务用途描述Develop run with tuning
模型适用产品/客户描述(人工填写)
模型重开发原因(如有)—
模型拟监控频率(人工填写)
模型路线LightGBM
目标变量gbflag
实验 IDDEV-mvyzwkut-49fe
待补充:模型开发/验证人员、应用频率、产品描述等字段需人工补填。

01 模型设计

建模数据样本窗口

项目内容
切分方式外部 data_flag 划分(TRAIN 内 6/4 随机开发/测试,VALI 跨时间,PSI 时点)
开发样本(Train)窗口2021-10-17 ~ 2026-04-30
预留验证样本窗口2024-04-30 ~ 2026-04-30
跨时间验证样本窗口2021-12-31 ~ 2026-06-30
PSI时点样本窗口2023-11-18 ~ 2026-07-31

模型好坏定义

项目内容
目标变量gbflag
坏样本(BAD=1)占比8.03%
好样本(BAD=0)占比—
不确定样本本流程要求目标变量已预计算为 0/1 二值,不确定样本请在上游合并口径

样本排除规则

序号排除规则说明
1id_no主键/时间/口径字段,不入模
2create_time主键/时间/口径字段,不入模
3data_flag主键/时间/口径字段,不入模
4gbflag 为空排除 0 行

滚动率分析

待补充:滚动率分析(M0→M6+ 状态迁移矩阵)需要逾期状态口径数据,当前流程目标变量为上游已计算的二值标签,暂未生成。
待补充:表现窗口及坏账率分析、模型细分定义需结合业务口径人工补充。

02 数据流

注:中文字典已加载(dictionary.csv,匹配 1 个字段);以下字段未在字典中、报告显示英文原名,请补充字典:ps_perd_no、gbflag、data_flag、id_no、age、max_int_rat、min_int_rat、min_tnr、sum_apply_amt、min_apply_amt、min_apprv_amt、apprv_amt_day、end_loan_to_rep_loan_freq、end_loan_to_cur_freq、f_rep_loan_to_f_apply_dt_freq、ps_prcp_amt_mon、ps_int_amt_mon、ps_int_amt_sum、ps_prcp_amt_avg、ps_int_amt_avg、setl_prcp_mon_ratio、acc_ps_perd_no、max_od_days、m_s_ps_instm_amt、f_ps_od_days_max、s_ps_od_days_max、t_ps_od_days_max、ovd_days_d3_num_m1、orig_prcp_avg_m3、orig_prcp_avg_m12 等共 127 个。

开发时点 — 样本排除

类别样本数
全部样本115682
排除:目标变量为空0
排除后样本115682

注:按照01模型设计的排除规则统计;当前流程的排除项为目标变量缺失。

总体样本概况

数据集样本量好样本坏样本坏样本率占比
全部样本11568210638992938.03%100.00%
开发样本380883493731518.27%32.92%
测试样本253912329121008.27%21.95%
训练+测试样本634795822852518.27%54.87%
跨时间验证样本347303229824327.00%30.02%
PSI时点样本17473158631610—15.10%

开发/预留验证样本交叉统计

类别开发+预留样本数开发样本数开发占比预留验证样本数预留验证占比
好582283493760.00%2329140.00%
坏5251315160.01%210039.99%
合计634793808860.00%2539140.00%

细致变量分析(预分箱口径,开发样本)

变量IV缺失率取值集中度最大相关系数
last_od_to_cur_freq0.19280.02%70.20%0.4978
max_od_days0.18700.02%25.81%0.5557
last_loan_setl_tnr0.16690.02%18.96%0.3034
end_loan_to_cur_freq0.14950.02%6.32%0.6425
apprv_amt_day0.14300.02%79.67%0.6396
cre_l60_used_limit_rate_min0.12720.06%10.21%0.7274
ps_perd_no0.12440.00%13.10%0.6425
loan_os_prcp_credit_amount_m30.12220.07%9.66%0.5204
cre_l90_used_limit_amt_min0.11600.06%14.62%0.7390
shd_pre_repay_due_rate0.11041.11%16.57%0.5562
odgt0_m12_subcurmonth0.10940.07%34.86%0.7415
cre_now_l60_used_amt_m_min0.10490.06%74.02%0.6396
cre_now_l60_limit_left_d_max0.10040.48%67.92%0.7660
loan_os_prcp_credit_amount_m60.09440.07%2.98%0.5204
f_ps_od_days_max0.09320.02%53.63%0.5202
suc_apprv_num_m10.09090.02%79.23%0.6800
acc_ps_perd_no0.08750.02%25.81%0.4281
last_od_num_m60.08590.02%78.72%0.6476
od_amt_m6_loan_os_prcp_m6_pc0.07980.07%93.14%0.6644
loan_os_prcp_credit_amount_m120.07950.07%0.23%0.2371
ps_int_amt_mon0.07750.02%5.50%0.7503
cre_l30_l90_used_amt_d_max0.07720.34%21.39%0.6593
available_limit_amt0.07710.02%32.48%0.8089
setl_prcp_ratio0.07680.02%21.83%0.6593
orig_prcp_avg_m30.07630.02%60.04%0.5989
cre_l30_l90_used_amt_m_max0.07310.06%21.48%0.6810
cre_now_l30_used_amt_d_min0.07105.71%86.91%0.9733
shd_max_dpd0.06901.05%74.62%0.3801
cre_l60_used_limit_rate_max0.06900.06%6.42%0.7274
od_amt_m12_loan_os_prcp_m12_pc0.06490.07%92.39%0.6644
repay_amount_m3_loan_os_prcp_pc0.06340.07%38.42%0.6366
t_ps_od_days_max0.06070.02%54.18%0.5455
apprv_num_m10.05950.02%73.93%0.7533
s_ps_od_days_max0.05940.02%54.12%0.5455
shd_l30_max_dpd0.05762.96%71.65%0.7038
apprv_num_m30.05720.02%52.89%0.7533
useratemore50maxcontimonsm120.05300.07%10.71%0.6543
cre_l30_limit_left_amt_min0.05040.06%4.18%0.7875
ps_int_amt_sum0.04650.02%1.21%0.7396
cre_now_l60_limit_left_d_min0.04565.74%11.24%0.8178
avg_tnr
平均履约月数
0.04560.02%30.34%0.6841
odgt0_m12_mob_pc0.04470.07%33.87%0.6158
cre_now_l60_used_amt_d_min0.042910.25%81.68%0.7521
max_loan_os_prcp_m120.04110.07%4.04%0.7503
odgt0_m12_num0.04070.07%34.86%0.7141
useratemore70_m6_num0.04050.07%28.42%0.6941
cre_now_l60_used_amt_m_max0.04000.06%11.21%0.6810
suc_apprv_num_m60.03950.02%34.85%0.7214
cre_now_p1m_used_rate_d0.03915.64%4.54%0.9733
useratemore70monthsubcurmonth0.03840.07%35.85%0.6077
orig_prcp_avg_m120.03720.02%3.24%0.6695
cre_l30_l90_used_amt_m_min0.03690.06%78.40%0.5989
shd_pre_repay_amt_rate0.03380.08%1.29%0.6789
cre_usedrate_u30_daydelta_max0.033012.49%83.72%0.6366
repay_amount_m6_all_shd_amt_pc0.03210.07%3.66%0.7885
ps_int_amt_avg0.03100.02%1.21%0.6095
cre_now_l30_used_amt_d_max0.02990.52%14.27%0.6000
cre_now_l30_limit_left_d_min0.02894.23%14.95%0.3058
repay_amount_m6_loan_os_prcp_pc0.02820.07%54.31%0.5912
shd_l60_min_dpd0.02782.50%51.07%0.7466
min_apply_amt0.02730.02%10.57%0.4938
useratemore90maxcontimonsm120.02690.07%24.43%0.6895
cre_now_l90_limit_left_d_min0.02556.91%9.77%0.8178
repay_amount_all_shd_amt_pc0.02540.07%5.67%0.6536
shd_instm_amt_sum0.02480.08%0.03%0.7396
useratemore90_m3_num0.02400.07%61.13%0.6941
cre_l30_l90_usedrate_m_max0.02280.06%26.85%0.7107
odgt0_m3_num0.02270.07%53.85%0.7141
cre_now_l60_usedrate_m_max0.02260.06%13.38%0.7107
m_s_ps_instm_amt0.02160.02%70.25%0.4301
min_tnr0.02150.02%53.18%0.6841
cre_now_p3m_used_rate_d0.021211.82%1.60%0.7521
setl_rem_prcp_paym_last_month_lo0.02030.07%87.85%0.6536
shd_l180_instm_amt_sum0.02000.08%0.11%0.7458
last_clear_num_m30.02000.02%57.66%0.7214
loan_os_prcp_gt0_m12_num_pc0.01960.07%61.32%0.5912
ps_prcp_amt_mon0.01830.02%4.60%0.7458
cre_now_p1m_used_rate_m0.01740.93%4.84%0.7553
cre_now_p2m_used_rate_m0.01581.66%2.23%0.7553
shd_l30_pre_repay_due_rate0.01533.11%48.56%0.7855
end_loan_to_rep_loan_freq0.01330.02%6.97%0.4280
shd_pre_repay_cnt0.01330.08%1.22%0.5711
shd_remain_actual_cnt0.01250.08%7.33%0.6008
account_age0.01240.07%4.74%0.7206
shd_l90_ps_due_cnt0.01180.08%17.68%0.6782
cre_now_l30_limit_left_d_max0.01160.72%79.30%0.7660
min_apprv_amt0.01100.02%29.17%0.5699
last_clear_num_m10.01000.02%79.20%0.7141
shd_min_dpd0.00941.06%26.93%0.7113
sum_apply_amt0.00870.02%10.45%0.4938
account_age_loan_day0.00840.07%2.11%0.7206
shd_remain_actual_rate0.00820.41%66.46%0.7855
max_int_rat0.00800.02%34.15%0.4054
last_apply_freq0.00800.02%87.14%0.6518
max_od_due0.00690.07%74.02%0.7415
setl_prcp_mon_ratio0.00610.07%6.05%0.4346
grade_limit_amount0.00510.02%38.22%0.4065
shd_l180_max_dpd0.00431.69%57.75%0.7885
cre_usedrate_u30_daydelta_min0.004012.49%81.81%0.4730
age0.00400.02%4.89%0.1655

注:共 124 个候选变量,按 IV 降序展示前 100 个;集中度 = 最大频次取值样本数 ÷ 非空样本数;相关性为皮尔逊(仅数值变量),完整数据见 variable_detail.json。

03 单变量分析(细分箱)

max_od_days — IV 0.2818,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]98278680114711.67%-0.3821
2(0.5, 5.5]215682048210865.04%0.5311
3(5.5, 7.5]231721142038.76%-0.0628
4(7.5, 12.5]2223194827512.37%-0.4482
5(12.5, inf]2145170643920.47%-1.0485
last_od_to_cur_freq — IV 0.2596,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 1.5]4011321679519.82%-1.0084
2(1.5, 3.5]260823722369.05%-0.0983
3(3.5, 8.5]20511970813.95%0.7854
4(8.5, 50042.0]26782607712.65%1.1973
5(50042.0, inf]267322476519677.36%0.1270
odgt0_m12_subcurmonth — IV 0.2143,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]98048659114511.68%-0.3825
2(0.5, 2.5]794975344155.22%0.4932
3(2.5, 5.5]402238601624.03%0.7651
4(5.5, 50005.0]30172932852.82%1.1351
5(50005.0, inf]1326811926134210.11%-0.2212
f_ps_od_days_max — IV 0.1937,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]2042218144227811.15%-0.3309
2(0.5, 1.5]844180024395.20%0.4970
3(1.5, 2.5]342032751454.24%0.7114
4(2.5, 4.5]359034531373.82%0.8211
5(4.5, inf]220720561516.84%0.2053
end_loan_to_cur_freq — IV 0.1904,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, -95.5]2062418471215310.44%-0.2566
2(-95.5, -68.5]301127622498.27%0.0003
3(-68.5, -22.5]632459293956.25%0.3028
4(-22.5, -0.5]571355651482.59%1.2211
5(-0.5, inf]240822032058.51%-0.0314
ovd_days_d3_num_m1 — IV 0.1789,2 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]346763222824487.06%0.1716
2(0.5, inf]3404270270220.62%-1.0581
last_loan_setl_tnr — IV 0.1645,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]723969942453.38%0.9456
2(0.5, 3.5]11453105429117.95%0.0426
3(3.5, 4.5]2791244234912.50%-0.4604
4(4.5, 8.5]106849521116310.89%-0.3034
5(8.5, inf]591354314828.15%0.0160
suc_apprv_num_m1 — IV 0.1594,3 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]301692728728829.55%-0.1580
2(0.5, 1.5]478745911964.09%0.7478
3(1.5, inf]31243052722.30%1.3409
apprv_amt_day — IV 0.1569,4 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 11.0]303442745228929.53%-0.1555
2(11.0, 1000.5]25062421853.39%0.9433
3(1000.5, 5995.0]33043220842.54%1.2404
4(5995.0, inf]19261837894.62%0.6213
t_ps_od_days_max — IV 0.1496,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]2063318517211610.26%-0.2368
2(0.5, 2.5]11097105245735.16%0.5046
3(2.5, 3.5]20481977713.47%0.9207
4(3.5, 5.5]219720751225.55%0.4278
5(5.5, inf]2105183726812.73%-0.4810
cre_now_l60_used_amt_m_min — IV 0.1485,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 1.7799999713897705]282012545827439.73%-0.1775
2(1.7799999713897705, 869.5950012207031]247623681084.36%0.6822
3(869.5950012207031, 1999.7100219726562]20602009512.48%1.2681
4(1999.7100219726562, 6462.93505859375]312530091163.71%0.8503
5(6462.93505859375, inf]220320711325.99%0.3475
cre_l60_used_limit_rate_min — IV 0.1478,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.019700000062584877]444143061353.04%1.0570
2(0.019700000062584877, 0.09174999967217445]281326561575.58%0.4228
3(0.09174999967217445, 0.2787500023841858]732368075167.05%0.1741
4(0.2787500023841858, 0.9161500036716461]188991715717429.22%-0.1181
5(0.9161500036716461, inf]4589398960013.07%-0.5111
ps_perd_no — IV 0.1475,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 1.5]499148451462.93%1.0963
2(1.5, 2.5]317029961745.49%0.4402
3(2.5, 3.5]322530072186.76%0.2184
4(3.5, 9.5]1737515601177410.21%-0.2317
5(9.5, inf]932784888399.00%-0.0916
orig_prcp_avg_m3 — IV 0.1453,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 23.449999809265137]2286420513235110.28%-0.2397
2(23.449999809265137, 1920.6666259765625]782875103184.06%0.7560
3(1920.6666259765625, 3515.75]293827901485.04%0.5306
4(3515.75, 6174.75]198018531276.41%0.2744
5(6174.75, inf]247022642068.34%-0.0089
s_ps_od_days_max — IV 0.1336,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]2061018430218010.58%-0.2713
2(0.5, 1.5]795175284235.32%0.4731
3(1.5, 2.5]346533031624.68%0.6090
4(2.5, 4.5]349433261684.81%0.5796
5(4.5, inf]256023432178.48%-0.0267
cre_l90_used_limit_amt_min — IV 0.1336,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 21.640000343322754]561854082103.74%0.8430
2(21.640000343322754, 748.7099914550781]807675265506.81%0.2107
3(748.7099914550781, 8516.22998046875]201721835418189.01%-0.0934
4(8516.22998046875, 14597.384765625]2267199027712.22%-0.4336
5(14597.384765625, inf]1932163729515.27%-0.6919
loan_os_prcp_credit_amount_m3 — IV 0.1276,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, -4999.0]3677329937810.28%-0.2392
2(-4999.0, 35.125]831579333824.59%0.6276
3(35.125, 100.47500228881836]191161761015067.88%0.0533
4(100.47500228881836, 124.34499740600586]2717241230511.23%-0.3378
5(124.34499740600586, inf]4235365757813.65%-0.5609
setl_prcp_ratio — IV 0.1275,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.035099999979138374]832579603654.38%0.6763
2(0.035099999979138374, 0.1652499958872795]448840913978.85%-0.0733
3(0.1652499958872795, 0.3135499954223633]4253372752612.37%-0.4479
4(0.3135499954223633, 0.6564500033855438]1207710849122810.17%-0.2273
5(0.6564500033855438, inf]893783036347.09%0.1664
apprv_num_m1 — IV 0.1225,4 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]281522542527279.69%-0.1734
2(0.5, 1.5]535051072434.54%0.6394
3(1.5, 2.5]19891914753.77%0.8335
4(2.5, inf]258924841054.06%0.7577
cre_now_l60_limit_left_d_max — IV 0.1208,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, -0.013299999758601189]2253199925411.27%-0.3455
2(-0.013299999758601189, 0.3223000019788742]393537202155.46%0.4423
3(0.3223000019788742, 0.6971500217914581]239722881094.55%0.6356
4(0.6971500217914581, 0.9999499917030334]28902809812.80%1.1376
5(0.9999499917030334, inf]264292396024699.34%-0.1360
cre_now_l30_limit_left_d_max — IV 0.1167,4 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.0002500000045984052]196918101598.08%0.0266
2(0.0002500000045984052, 0.42524999380111694]220321031004.54%0.6404
3(0.42524999380111694, 0.9999499917030334]30422974682.24%1.3726
4(0.9999499917030334, inf]306002779828029.16%-0.1110
shd_pre_repay_due_rate — IV 0.1142,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.868800014257431]1047399205535.28%0.4846
2(0.868800014257431, 0.9498499929904938]654960524977.59%0.0972
3(0.9498499929904938, 0.9933499991893768]563651215159.14%-0.1054
4(0.9933499991893768, 1.0524500012397766]95728408116412.16%-0.4250
5(1.0524500012397766, inf]543450373977.31%0.1383
last_od_num_m6 — IV 0.1099,4 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]299772791920586.87%0.2016
2(0.5, 3.5]4179362155813.35%-0.5358
3(3.5, 7.5]1999171028914.46%-0.6281
4(7.5, inf]1925168024512.73%-0.4807
available_limit_amt — IV 0.1061,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.03499999921768904]1238510918146711.85%-0.3987
2(0.03499999921768904, 468.864990234375]452342472766.10%0.3276
3(468.864990234375, 4307.300048828125]11998111488507.08%0.1678
4(4307.300048828125, 27316.265625]725067884626.37%0.2814
5(27316.265625, inf]19241829954.94%0.5517
loan_os_prcp_credit_amount_m6 — IV 0.1031,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 4.129999876022339]191717931246.47%0.2656
2(4.129999876022339, 33.795000076293945]621059412694.33%0.6892
3(33.795000076293945, 48.94499969482422]443741412966.67%0.2326
4(48.94499969482422, 107.59499740600586]187351707416618.87%-0.0756
5(107.59499740600586, inf]6761596279911.82%-0.3959
cre_now_l30_used_amt_d_min — IV 0.1011,3 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 1.0000500082969666]312112827329389.41%-0.0980
2(1.0000500082969666, 1.382200002670288]19331860733.78%0.8757
3(1.382200002670288, inf]27682687812.93%1.1395
cre_l30_l90_used_amt_d_max — IV 0.0993,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.508650004863739]289127071846.36%0.2851
2(0.508650004863739, 0.6789500117301941]800773646438.03%0.0347
3(0.6789500117301941, 0.8662499785423279]1519013563162710.71%-0.2829
4(0.8662499785423279, 0.9735499918460846]308228212618.47%-0.0232
5(0.9735499918460846, inf]879083584324.91%0.5590
acc_ps_perd_no — IV 0.0987,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]98278680114711.67%-0.3821
2(0.5, 7.5]132211211111108.40%-0.0162
3(7.5, 23.5]846178965656.68%0.2313
4(23.5, 65.5]462543732525.45%0.4478
5(65.5, inf]19461870763.91%0.7970
apprv_num_m3 — IV 0.0970,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]2013918040209910.42%-0.2548
2(0.5, 1.5]663762224156.25%0.3016
3(1.5, 2.5]356333532105.89%0.3646
4(2.5, 4.5]361034311794.96%0.5473
5(4.5, inf]413138842475.98%0.3493
ps_int_amt_mon — IV 0.0946,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 1.475000023841858]23542275793.36%0.9543
2(1.475000023841858, 19.704999923706055]933887056336.78%0.2152
3(19.704999923706055, 132.375]194131782715868.17%0.0136
4(132.375, 296.6699981689453]4803426753611.16%-0.3314
5(296.6699981689453, inf]2172185631614.55%-0.6355
loan_os_prcp_credit_amount_m12 — IV 0.0894,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 16.18000030517578]316430581063.35%0.9564
2(16.18000030517578, 44.21000099182129]603356713626.00%0.3458
3(44.21000099182129, 77.875]144531322312308.51%-0.0308
4(77.875, 95.76499938964844]663560086279.45%-0.1458
5(95.76499938964844, inf]7775695182410.60%-0.2732
cre_l30_l90_used_amt_m_max — IV 0.0882,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 35.010000228881836]863382194144.80%0.5828
2(35.010000228881836, 350.3800048828125]631158194927.80%0.0649
3(350.3800048828125, 1554.1449584960938]132551202012359.32%-0.1300
4(1554.1449584960938, 1951.3800048828125]205918961637.92%0.0482
5(1951.3800048828125, inf]7807696184610.84%-0.2980
avg_od_amt_m6_orig_prcp_pc — IV 0.0815,2 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5699999928474426]355683286627027.60%0.0927
2(0.5699999928474426, inf]2492204544717.94%-0.8851
od_amt_m6_loan_os_prcp_m6_pc — IV 0.0804,2 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.38500000536441803]355693286527047.60%0.0920
2(0.38500000536441803, inf]2491204644517.86%-0.8801
cre_now_l60_used_amt_d_min — IV 0.0776,4 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 1.000950038433075]279432520527389.80%-0.1106
2(1.000950038433075, 1.4549500346183777]235622351215.14%0.5858
3(1.4549500346183777, 2.305649995803833]18601789713.82%0.8963
4(2.305649995803833, inf]202619261004.94%0.6276
cre_l60_used_limit_rate_max — IV 0.0767,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.22405000030994415]426540422235.23%0.4918
2(0.22405000030994415, 0.9925500154495239]205931889516988.25%0.0039
3(0.9925500154495239, 1.2138000130653381]592755623656.16%0.3183
4(1.2138000130653381, 1.424750030040741]2499215134813.93%-0.5840
5(1.424750030040741, inf]4781426551610.79%-0.2934
shd_max_dpd — IV 0.0690,2 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]95658427113811.90%-0.4008
2(0.5, inf]281222613419887.07%0.1731
loan_os_prcp_gt0_m3_num_pc — IV 0.0681,2 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 68.16500091552734]335432521023.04%1.0563
2(68.16500091552734, inf]347063165930478.78%-0.0649
repay_amount_m3_loan_os_prcp_pc — IV 0.0661,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, -4999.0]480445542505.20%0.4966
2(-4999.0, 31.585000038146973]5227462060711.61%-0.3761
3(31.585000038146973, 41.34499931335449]3887349639110.06%-0.2150
4(41.34499931335449, 74.81499862670898]943485768589.09%-0.1036
5(74.81499862670898, inf]147081366510437.09%0.1670
od_amt_m12_loan_os_prcp_m12_pc — IV 0.0658,2 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.0950000025331974]352223253226907.64%0.0870
2(0.0950000025331974, inf]2838237945916.17%-0.7603
max_od_due — IV 0.0638,2 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]98078661114611.69%-0.3832
2(0.5, inf]282532625020037.09%0.1673
cre_now_l90_limit_left_d_min — IV 0.0635,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, -0.6798500120639801]270025101907.04%0.1917
2(-0.6798500120639801, 0.1424499973654747]288626242629.08%-0.0852
3(0.1424499973654747, 0.9937500059604645]4103356354013.16%-0.5025
4(0.9937500059604645, 1.1872499585151672]791274344786.04%0.3549
5(1.1872499585151672, inf]178551634715088.45%-0.0061
cre_now_l60_limit_left_d_min — IV 0.0600,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.04739999957382679]363233442887.93%0.0650
2(0.04739999957382679, 0.9717499911785126]4407384256512.82%-0.4700
3(0.9717499911785126, 1.149049997329712]931387205936.37%0.3013
4(1.149049997329712, 3.562299966812134]152411389013518.86%-0.0566
5(3.562299966812134, inf]330830832256.80%0.2306
cre_l30_limit_left_amt_min — IV 0.0597,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, -132.05500030517578]7225636985611.85%-0.3986
2(-132.05500030517578, 101.06500244140625]428240322505.84%0.3750
3(101.06500244140625, 432.6999969482422]312328952287.30%0.1359
4(432.6999969482422, 2482.1400146484375]1019293188748.58%-0.0389
5(2482.1400146484375, inf]13243123019427.11%0.1639
useratemore50maxcontimonsm12 — IV 0.0578,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]407638891874.59%0.6291
2(0.5, 1.5]200318701336.64%0.2376
3(1.5, 2.5]228621651215.29%0.4787
4(2.5, 4.5]589354014928.35%-0.0099
5(4.5, inf]238022158622169.31%-0.1294
cre_now_p1m_used_rate_m — IV 0.0577,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, -0.13974999636411667]513748073306.42%0.2755
2(-0.13974999636411667, -4.999999873689376e-05]21512063884.09%0.7513
3(-4.999999873689376e-05, 0.0002500000045984052]1903165724612.93%-0.4959
4(0.0002500000045984052, 0.04154999926686287]429440022926.80%0.2145
5(0.04154999926686287, inf]242492207621738.96%-0.0849
shd_l30_max_dpd — IV 0.0576,2 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]278582582320357.30%0.1535
2(0.5, inf]91018026107511.81%-0.3769
cre_now_p1m_used_rate_d — IV 0.0563,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.6786499917507172]654861234256.49%0.3004
2(0.6786499917507172, 0.9951499998569489]218641984220229.25%-0.0836
3(0.9951499998569489, 1.0000500082969666]1839157826114.19%-0.5679
4(1.0000500082969666, 1.7942500114440918]380435272777.28%0.1769
5(1.7942500114440918, inf]18851790955.04%0.5688
ps_int_amt_sum — IV 0.0558,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 209.51499938964844]317330021715.39%0.4594
2(209.51499938964844, 815.2900085449219]1035696177397.14%0.1600
3(815.2900085449219, 1891.7449951171875]1042996138167.82%0.0605
4(1891.7449951171875, 10533.93994140625]113161027610409.19%-0.1154
5(10533.93994140625, inf]2806242238413.69%-0.5642
shd_l30_pre_repay_due_rate — IV 0.0546,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.8838499784469604]6969614482511.84%-0.3781
2(0.8838499784469604, 1.0121999979019165]179551658813677.61%0.1101
3(1.0121999979019165, 1.781749963760376]287526032729.46%-0.1274
4(1.781749963760376, 4.9175498485565186]709165455467.70%0.0979
5(4.9175498485565186, inf]20131914994.92%0.5758
cre_now_l60_used_amt_m_max — IV 0.0541,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 83.30500030517578]514849042444.74%0.5951
2(83.30500030517578, 1001.9599914550781]169371556813698.08%0.0256
3(1001.9599914550781, 2003.010009765625]715165156368.89%-0.0789
4(2003.010009765625, 2696.469970703125]2318205426411.39%-0.3539
5(2696.469970703125, inf]651158746379.78%-0.1840
cre_l30_l90_used_amt_m_min — IV 0.0527,4 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.6299999952316284]298562718526718.95%-0.0853
2(0.6299999952316284, 1107.5999755859375]285827421164.06%0.7573
3(1107.5999755859375, 5082.195068359375]325830821765.40%0.4573
4(5082.195068359375, inf]209319061878.93%-0.0839
avg_tnr 平均履约月数 — IV 0.0524,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 3.5]24362340963.94%0.7876
2(3.5, 5.5]367234422306.26%0.2998
3(5.5, 6.5]266024721887.07%0.1704
4(6.5, 10.5]126711164510268.10%0.0233
5(10.5, inf]166411503116109.67%-0.1721
cre_now_l30_limit_left_d_min — IV 0.0482,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.26759999990463257]261323772369.03%-0.0819
2(0.26759999990463257, 0.997050017118454]3523306645712.97%-0.4882
3(0.997050017118454, 1.1242499947547913]137731277010037.28%0.1524
4(1.1242499947547913, 1.7257500290870667]116271058510428.96%-0.0734
5(1.7257500290870667, inf]493946203196.46%0.2813
max_loan_os_prcp_m12 — IV 0.0469,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 1747.2849731445312]410138602415.88%0.3679
2(1747.2849731445312, 2010.25]280625622448.70%-0.0543
3(2010.25, 13572.5849609375]224802076617147.62%0.0888
4(13572.5849609375, 20848.58984375]361932783419.42%-0.1426
5(20848.58984375, inf]5054444560912.05%-0.4180
cre_usedrate_u30_daydelta_max — IV 0.0468,3 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 57.5]282927001294.56%0.6900
2(57.5, 85.5]211219921205.68%0.4582
3(85.5, inf]283912574126509.33%-0.0776
odgt0_m12_num — IV 0.0464,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]1326811926134210.11%-0.2212
2(0.5, 1.5]773771006378.23%0.0054
3(1.5, 2.5]553450984367.88%0.0532
4(2.5, 5.5]832177475746.90%0.1967
5(5.5, inf]320030401605.00%0.5387
odgt0_m12_mob_pc — IV 0.0450,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 4.164999961853027]1327211930134210.11%-0.2208
2(4.164999961853027, 20.130000114440918]125321151210208.14%0.0179
3(20.130000114440918, 37.14500045776367]653261104226.46%0.2670
4(37.14500045776367, 57.670000076293945]347132192527.26%0.1417
5(57.670000076293945, inf]225321401135.02%0.5355
suc_apprv_num_m6 — IV 0.0434,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]132721196013129.89%-0.1959
2(0.5, 1.5]873979657748.86%-0.0747
3(1.5, 2.5]477044383326.96%0.1869
4(2.5, 3.5]304228022407.89%0.0515
5(3.5, inf]825777654925.96%0.3530
useratemore70monthsubcurmonth — IV 0.0434,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]136451244412018.80%-0.0676
2(0.5, 2.5]658659736139.31%-0.1291
3(2.5, 4.5]519846855139.87%-0.1939
4(4.5, 9.5]603755644737.83%0.0593
5(9.5, inf]659462453495.29%0.4787
setl_rem_prcp_paym_m3_num — IV 0.0433,2 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]320942924028548.89%-0.0789
2(0.5, inf]596656712954.94%0.5504
orig_prcp_avg_m12 — IV 0.0431,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 1031.9722290039062]11498107357636.64%0.2381
2(1031.9722290039062, 1871.875]692564055207.51%0.1051
3(1871.875, 3819.857177734375]872779937348.41%-0.0181
4(3819.857177734375, 13787.5]890980418689.74%-0.1798
5(13787.5, inf]2021175626513.11%-0.5149
useratemore70_m6_num — IV 0.0429,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]10817101466716.20%0.3103
2(0.5, 1.5]443541153207.22%0.1484
3(1.5, 4.5]144121310413089.08%-0.1013
4(4.5, 5.5]346831323369.69%-0.1734
5(5.5, inf]4928441451410.43%-0.2554
shd_pre_repay_amt_rate — IV 0.0409,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.6238499879837036]12637118148236.51%0.2584
2(0.6238499879837036, 0.7506499886512756]863179446877.96%0.0421
3(0.7506499886512756, 0.8537499904632568]906682188489.35%-0.1345
4(0.8537499904632568, 0.9255499839782715]5759513162810.90%-0.3052
5(0.9255499839782715, inf]196618031638.29%-0.0022
repay_amount_m6_all_shd_amt_pc — IV 0.0386,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 93.98500061035156]132571194613119.89%-0.1961
2(93.98500061035156, 95.59499740600586]303328122217.29%0.1378
3(95.59499740600586, 96.48500061035156]297526942819.45%-0.1453
4(96.48500061035156, 99.48500061035156]137351268710487.63%0.0880
5(99.48500061035156, inf]506047722885.69%0.4018
m_s_ps_instm_amt — IV 0.0386,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 33.55999946594238]268902491619747.34%0.1295
2(33.55999946594238, 371.93499755859375]5271473953210.09%-0.2190
3(371.93499755859375, 614.2250061035156]2004177423011.48%-0.3630
4(614.2250061035156, 1121.7000122070312]196017851758.93%-0.0836
5(1121.7000122070312, inf]1955171623912.23%-0.4347
shd_l60_min_dpd — IV 0.0368,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, -286.5]11318103389808.66%-0.0354
2(-286.5, -151.5]205219031497.26%0.1558
3(-151.5, -0.5]18831799844.46%0.6727
4(-0.5, 0.5]189651742415418.13%0.0340
5(0.5, inf]2917255835912.31%-0.4278
cre_now_l30_used_amt_d_max — IV 0.0362,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.6786499917507172]737668984786.48%0.2678
2(0.6786499917507172, 0.7513499855995178]240421862189.07%-0.0962
3(0.7513499855995178, 0.7632499933242798]191617791377.15%0.1623
4(0.7632499933242798, 0.9139499962329865]174711577117009.73%-0.1740
5(0.9139499962329865, inf]872481106147.04%0.1793
ps_int_amt_avg — IV 0.0360,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 4.333299875259399]408338572265.54%0.4312
2(4.333299875259399, 13.080849647521973]12227113149137.47%0.1111
3(13.080849647521973, 15.02239990234375]204518412049.98%-0.2060
4(15.02239990234375, 38.24515151977539]123061126510418.46%-0.0244
5(38.24515151977539, inf]7419665376610.32%-0.2443
useratemore90maxcontimonsm12 — IV 0.0350,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]929887235756.18%0.3136
2(0.5, 1.5]690363495548.03%0.0332
3(1.5, 4.5]147501346112898.74%-0.0598
4(4.5, 7.5]465241884649.97%-0.2056
5(7.5, inf]2457219026710.87%-0.3013
shd_l90_ps_due_cnt — IV 0.0336,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 2.5]248623831034.14%0.7357
2(2.5, 5.5]935585747818.35%-0.0098
3(5.5, 9.5]879379678269.39%-0.1392
4(9.5, 17.5]708265465367.57%0.0968
5(17.5, inf]1034394409038.73%-0.0587
repay_amount_all_shd_amt_pc — IV 0.0306,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, -4999.0]215920171426.58%0.2478
2(-4999.0, 90.52499771118164]920783069019.79%-0.1845
3(90.52499771118164, 97.21500015258789]160891468914008.70%-0.0551
4(97.21500015258789, 98.58499908447266]698264825007.16%0.1565
5(98.58499908447266, inf]362334172065.69%0.4029
min_apply_amt — IV 0.0305,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 1135.0]3388301037811.16%-0.3311
2(1135.0, 2270.0]514247344087.93%0.0453
3(2270.0, 4180.0]128411167911629.05%-0.0983
4(4180.0, 7300.0]922184907317.93%0.0463
5(7300.0, inf]748870174716.29%0.2953
last_clear_num_m3 — IV 0.0305,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]219561990820489.33%-0.1317
2(0.5, 1.5]704365255187.35%0.1275
3(1.5, 2.5]341731772407.02%0.1771
4(2.5, 3.5]203219031296.35%0.2854
5(3.5, inf]363234172155.92%0.3599
useratemore90_m3_num — IV 0.0302,4 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]232662150117657.59%0.0942
2(0.5, 1.5]611756464717.70%0.0781
3(1.5, 2.5]397936233568.95%-0.0856
4(2.5, inf]4698414155711.86%-0.3996
shd_instm_amt_sum — IV 0.0298,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 14048.51513671875]12742117869567.50%0.1062
2(14048.51513671875, 17727.4052734375]271225471656.08%0.3310
3(17727.4052734375, 134568.0234375]178261634114858.33%-0.0074
4(134568.0234375, 248682.5859375]2691240928210.48%-0.2606
5(248682.5859375, inf]2088182726112.50%-0.4598
cre_now_l60_usedrate_m_max — IV 0.0295,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 4.999999873689376e-05]509447493456.77%0.2166
2(4.999999873689376e-05, 0.039750000461936]263224991335.05%0.5278
3(0.039750000461936, 0.08044999837875366]307928462337.57%0.0971
4(0.08044999837875366, 0.5794999897480011]244692224522249.09%-0.1027
5(0.5794999897480011, inf]279125762157.70%0.0778
repay_amount_m6_loan_os_prcp_pc — IV 0.0294,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, -4999.0]682863534756.96%0.1877
2(-4999.0, 51.885000228881836]1923168224112.53%-0.4628
3(51.885000228881836, 61.88999938964844]2837253829910.54%-0.2670
4(61.88999938964844, 93.5250015258789]578952605299.14%-0.1088
5(93.5250015258789, inf]206831907816057.76%0.0697
min_tnr — IV 0.0283,4 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 1.5]268824342549.45%-0.1460
2(1.5, 5.0]202611882614357.08%0.1681
3(5.0, 8.5]301227372759.13%-0.1081
4(8.5, inf]121191093311869.79%-0.1847
ps_prcp_amt_mon — IV 0.0282,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 38.48500061035156]206919661034.98%0.5431
2(38.48500061035156, 955.4700012207031]225732080117727.85%0.0569
3(955.4700012207031, 2836.0150146484375]911382578569.39%-0.1394
4(2836.0150146484375, 4656.4951171875]226720851828.03%0.0326
5(4656.4951171875, inf]2058182123711.52%-0.3669
cre_now_p3m_used_rate_d — IV 0.0274,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.40674999356269836]436340892746.28%0.3566
2(0.40674999356269836, 0.5166500210762024]454141643778.30%0.0557
3(0.5166500210762024, 0.8346999883651733]128661159012769.92%-0.1399
4(0.8346999883651733, 1.7775499820709229]847277047689.07%-0.0406
5(1.7775499820709229, inf]334331042397.15%0.2177
cre_now_p2m_used_rate_m — IV 0.0265,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, -0.11755000054836273]793973805597.04%0.1785
2(-0.11755000054836273, -0.00014999999621068127]239722681295.38%0.4649
3(-0.00014999999621068127, 0.0799499973654747]533549014348.13%0.0223
4(0.0799499973654747, 0.35944999754428864]170241542615989.39%-0.1346
5(0.35944999754428864, inf]476243723908.19%0.0149
cre_l30_l90_usedrate_m_max — IV 0.0258,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.025150000117719173]11879110807996.73%0.2240
2(0.025150000117719173, 0.0799499973654747]331030532577.76%0.0693
3(0.0799499973654747, 0.1587500050663948]6612594866410.04%-0.2130
4(0.1587500050663948, 0.5746999979019165]143061303312738.90%-0.0794
5(0.5746999979019165, inf]195818011578.02%0.0343
loan_os_prcp_gt0_m12_num_pc — IV 0.0249,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 57.73500061035156]329831381604.85%0.5705
2(57.73500061035156, 72.08000183105469]263124172148.13%0.0186
3(72.08000183105469, 76.01499938964844]196117841779.03%-0.0953
4(76.01499938964844, 92.31500244140625]664060945468.22%0.0067
5(92.31500244140625, inf]235302147820528.72%-0.0575
shd_l180_instm_amt_sum — IV 0.0243,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 5366.869873046875]147451367710687.24%0.1442
2(5366.869873046875, 6463.929931640625]205118541979.61%-0.1638
3(6463.929931640625, 34552.490234375]166591528313768.26%0.0019
4(34552.490234375, 55994.666015625]2458220725110.21%-0.2318
5(55994.666015625, inf]2146188925711.98%-0.4110
odgt0_m3_num — IV 0.0227,4 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]204941889715977.79%0.0652
2(0.5, 1.5]956088227387.72%0.0753
3(1.5, 2.5]536648824849.02%-0.0945
4(2.5, inf]2640231033012.50%-0.4598
grade_limit_amount — IV 0.0226,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 84500.0]2178194523310.70%-0.2840
2(84500.0, 115500.0]941686567608.07%0.0267
3(115500.0, 137500.0]203319001336.54%0.2533
4(137500.0, 180500.0]6132550163110.29%-0.2406
5(180500.0, inf]183211692813937.60%0.0916
end_loan_to_rep_loan_freq — IV 0.0224,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 1003.5]286772641922587.87%0.0537
2(1003.5, 1935.5]201218191939.59%-0.1626
3(1935.5, 2132.5]2266198428212.44%-0.4550
4(2132.5, 2322.5]266624172499.34%-0.1331
5(2322.5, inf]245922911686.83%0.2068
shd_pre_repay_cnt — IV 0.0209,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 25.5]956088766847.15%0.1575
2(25.5, 91.5]11526105489788.49%-0.0275
3(91.5, 237.5]753169775547.36%0.1275
4(237.5, 362.5]253723142238.79%-0.0661
5(362.5, inf]6905619571010.28%-0.2395
tothismonthodcontireducenumm12 — IV 0.0209,2 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]328993004428558.68%-0.0521
2(0.5, inf]516148672945.70%0.4009
last_clear_num_m1 — IV 0.0203,3 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]301612750926528.79%-0.0667
2(0.5, 1.5]491645943226.55%0.2520
3(1.5, inf]300328271765.86%0.3705
setl_rem_prcp_paym_last_month_lo — IV 0.0202,2 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, -4999.0]289927571424.90%0.5604
2(-4999.0, inf]351613215430078.55%-0.0361
tothismonthodcontiaddnumm12 — IV 0.0187,2 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]330573046325947.85%0.0576
2(0.5, inf]5003444855511.09%-0.3245
shd_remain_actual_cnt — IV 0.0182,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 2.5]433940353047.01%0.1800
2(2.5, 3.5]279026271635.84%0.3742
3(3.5, 5.5]413137923398.21%0.0090
4(5.5, 35.5]218761990819689.00%-0.0916
5(35.5, inf]492345483757.62%0.0898
last_apply_freq — IV 0.0173,3 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 6.5]333023067926237.88%0.0533
2(6.5, 498.5]2175192425111.54%-0.3692
3(498.5, inf]2603232727610.60%-0.2740
max_int_rat — IV 0.0172,4 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.2142000049352646]372734912366.33%0.2882
2(0.2142000049352646, 0.2332499995827675]1080998809298.59%-0.0418
3(0.2332499995827675, 0.23605000227689743]1006891099599.53%-0.1548
4(0.23605000227689743, inf]134761245010267.61%0.0901
account_age — IV 0.0164,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 21.5]178221650113217.41%0.1193
2(21.5, 24.5]323129502818.70%-0.0545
3(24.5, 29.5]428339473367.85%0.0579
4(29.5, 81.5]996589939729.75%-0.1809
5(81.5, inf]275925202398.66%-0.0502
shd_perd_max — IV 0.0161,2 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 8.5]20471951964.69%0.6061
2(8.5, inf]360123295930538.48%-0.0265
exposure — IV 0.0155,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 928.0249938964844]278625112759.87%-0.1943
2(928.0249938964844, 1019.9500122070312]352932982316.55%0.2527
3(1019.9500122070312, 1748.8150024414062]3229290032910.19%-0.2295
4(1748.8150024414062, 14424.9599609375]211701950116697.88%0.0523
5(14424.9599609375, inf]736667206468.77%-0.0639
setl_prcp_mon_ratio — IV 0.0149,5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.0011139739654026926]2888259229610.25%-0.2359
2(0.0011139739654026926, 0.017109270207583904]1010492288768.67%-0.0511
3(0.017109270207583904, 0.0826290100812912]188811738714947.91%0.0485
4(0.0826290100812912, 0.09751652926206589]383734913469.02%-0.0942
5(0.09751652926206589, inf]235022131375.83%0.3764

注:统计全部候选变量的自动分箱明细(初分组),按 IV 降序展示前 100 个变量,共 124 个;完整数据见 feature_report.json。

04 单变量分析(粗分箱)

f_ps_od_days_max — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]2042218144227811.15%-0.3309
2(0.5, 1.5]844180024395.20%0.4970
3(1.5, 2.5]342032751454.24%0.7114
4(2.5, 4.5]359034531373.82%0.8211
5(4.5, inf]220720561516.84%0.2053
2026-11-02T16:27:55.081606 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(11.15% → 11.15% → 6.84%),如需严格单调可考虑调整切点或合并分箱。IV=0.1937,该变量区分度尚可。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 4→5 箱有局部反弹:Bad Rate 由 3.82% 回升至 6.84%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 0.5])为坏率最高区间:坏率高达 11.15%,可作为拒绝或重点风控策略的阈值参考

suc_apprv_num_m1 — 3 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]301692728728829.55%-0.1580
2(0.5, 1.5]478745911964.09%0.7478
3(1.5, inf]31243052722.30%1.3409
2026-11-02T16:27:55.207670 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调下降(9.55% → 2.30%),WOE 单调上升(-0.1580 → 1.3409),符合风控分箱预期。IV=0.1594,该变量区分度尚可。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 2 箱 WOE<0(风险高于平均)

第 1 箱((-inf, 0.5])为坏率最高区间:坏率高达 9.55%,可作为拒绝或重点风控策略的阈值参考

apprv_amt_day — 4 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 11.0]303442745228929.53%-0.1555
2(11.0, 1000.5]25062421853.39%0.9433
3(1000.5, 5995.0]33043220842.54%1.2404
4(5995.0, inf]19261837894.62%0.6213
2026-11-02T16:27:55.381712 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(9.53% → 9.53% → 4.62%),如需严格单调可考虑调整切点或合并分箱。IV=0.1569,该变量区分度尚可。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 3→4 箱有局部反弹:Bad Rate 由 2.54% 回升至 4.62%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 11.0])为坏率最高区间:坏率高达 9.53%,可作为拒绝或重点风控策略的阈值参考

t_ps_od_days_max — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]2063318517211610.26%-0.2368
2(0.5, 2.5]11097105245735.16%0.5046
3(2.5, 3.5]20481977713.47%0.9207
4(3.5, 5.5]219720751225.55%0.4278
5(5.5, inf]2105183726812.73%-0.4810
2026-11-02T16:27:55.504027 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(10.26% → 12.73% → 12.73%),如需严格单调可考虑调整切点或合并分箱。IV=0.1496,该变量区分度尚可。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 1→2 箱有局部反弹:Bad Rate 由 10.26% 回落至 5.16%,如需严格单调可考虑合并这两箱

第 5 箱((5.5, inf])为坏率最高区间:坏率高达 12.73%,可作为拒绝或重点风控策略的阈值参考

cre_now_l60_used_amt_m_min — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 1.7799999713897705]282012545827439.73%-0.1775
2(1.7799999713897705, 869.5950012207031]247623681084.36%0.6822
3(869.5950012207031, 1999.7100219726562]20602009512.48%1.2681
4(1999.7100219726562, 6462.93505859375]312530091163.71%0.8503
5(6462.93505859375, inf]220320711325.99%0.3475
2026-11-02T16:27:55.634486 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(9.73% → 9.73% → 5.99%),如需严格单调可考虑调整切点或合并分箱。IV=0.1485,该变量区分度尚可。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 3→4 箱有局部反弹:Bad Rate 由 2.48% 回升至 3.71%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 1.7799999713897705])为坏率最高区间:坏率高达 9.73%,可作为拒绝或重点风控策略的阈值参考

cre_l60_used_limit_rate_min — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.019700000062584877]444143061353.04%1.0570
2(0.019700000062584877, 0.09174999967217445]281326561575.58%0.4228
3(0.09174999967217445, 0.2787500023841858]732368075167.05%0.1741
4(0.2787500023841858, 0.9161500036716461]188991715717429.22%-0.1181
5(0.9161500036716461, inf]4589398960013.07%-0.5111
2026-11-02T16:27:55.768517 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调上升(3.04% → 13.07%),WOE 单调下降(1.0570 → -0.5111),符合风控分箱预期。IV=0.1478,该变量区分度尚可。

WOE 在第 3、4 箱之间穿越 0 线:前 3 箱 WOE>0(客户质量优于整体),后 2 箱 WOE<0(风险高于平均)

第 5 箱((0.9161500036716461, inf])为坏率最高区间:坏率高达 13.07%,可作为拒绝或重点风控策略的阈值参考

s_ps_od_days_max — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]2061018430218010.58%-0.2713
2(0.5, 1.5]795175284235.32%0.4731
3(1.5, 2.5]346533031624.68%0.6090
4(2.5, 4.5]349433261684.81%0.5796
5(4.5, inf]256023432178.48%-0.0267
2026-11-02T16:27:55.894238 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(10.58% → 10.58% → 8.48%),如需严格单调可考虑调整切点或合并分箱。IV=0.1336,该变量区分度尚可。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 3→4 箱有局部反弹:Bad Rate 由 4.68% 回升至 4.81%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 0.5])为坏率最高区间:坏率高达 10.58%,可作为拒绝或重点风控策略的阈值参考

cre_l90_used_limit_amt_min — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 21.640000343322754]561854082103.74%0.8430
2(21.640000343322754, 748.7099914550781]807675265506.81%0.2107
3(748.7099914550781, 8516.22998046875]201721835418189.01%-0.0934
4(8516.22998046875, 14597.384765625]2267199027712.22%-0.4336
5(14597.384765625, inf]1932163729515.27%-0.6919
2026-11-02T16:27:56.023657 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调上升(3.74% → 15.27%),WOE 单调下降(0.8430 → -0.6919),符合风控分箱预期。IV=0.1336,该变量区分度尚可。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 5 箱((14597.384765625, inf])为坏率最高区间:坏率高达 15.27%,可作为拒绝或重点风控策略的阈值参考

loan_os_prcp_credit_amount_m3 — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, -4999.0]3677329937810.28%-0.2392
2(-4999.0, 35.125]831579333824.59%0.6276
3(35.125, 100.47500228881836]191161761015067.88%0.0533
4(100.47500228881836, 124.34499740600586]2717241230511.23%-0.3378
5(124.34499740600586, inf]4235365757813.65%-0.5609
2026-11-02T16:27:56.153554 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(10.28% → 13.65% → 13.65%),如需严格单调可考虑调整切点或合并分箱。IV=0.1276,该变量区分度尚可。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 1→2 箱有局部反弹:Bad Rate 由 10.28% 回落至 4.59%,如需严格单调可考虑合并这两箱

第 5 箱((124.34499740600586, inf])为坏率最高区间:坏率高达 13.65%,可作为拒绝或重点风控策略的阈值参考

apprv_num_m1 — 4 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]281522542527279.69%-0.1734
2(0.5, 1.5]535051072434.54%0.6394
3(1.5, 2.5]19891914753.77%0.8335
4(2.5, inf]258924841054.06%0.7577
2026-11-02T16:27:56.275508 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(9.69% → 9.69% → 4.06%),如需严格单调可考虑调整切点或合并分箱。IV=0.1225,该变量区分度尚可。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 3→4 箱有局部反弹:Bad Rate 由 3.77% 回升至 4.06%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 0.5])为坏率最高区间:坏率高达 9.69%,可作为拒绝或重点风控策略的阈值参考

cre_now_l60_limit_left_d_max — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, -0.013299999758601189]2253199925411.27%-0.3455
2(-0.013299999758601189, 0.3223000019788742]393537202155.46%0.4423
3(0.3223000019788742, 0.6971500217914581]239722881094.55%0.6356
4(0.6971500217914581, 0.9999499917030334]28902809812.80%1.1376
5(0.9999499917030334, inf]264292396024699.34%-0.1360
2026-11-02T16:27:56.396988 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(11.27% → 11.27% → 9.34%),如需严格单调可考虑调整切点或合并分箱。IV=0.1208,该变量区分度尚可。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 4→5 箱有局部反弹:Bad Rate 由 2.80% 回升至 9.34%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, -0.013299999758601189])为坏率最高区间:坏率高达 11.27%,可作为拒绝或重点风控策略的阈值参考

available_limit_amt — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.03499999921768904]1238510918146711.85%-0.3987
2(0.03499999921768904, 468.864990234375]452342472766.10%0.3276
3(468.864990234375, 4307.300048828125]11998111488507.08%0.1678
4(4307.300048828125, 27316.265625]725067884626.37%0.2814
5(27316.265625, inf]19241829954.94%0.5517
2026-11-02T16:27:56.526507 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(11.85% → 11.85% → 4.94%),如需严格单调可考虑调整切点或合并分箱。IV=0.1061,该变量区分度尚可。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 2→3 箱有局部反弹:Bad Rate 由 6.10% 回升至 7.08%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 0.03499999921768904])为坏率最高区间:坏率高达 11.85%,可作为拒绝或重点风控策略的阈值参考

loan_os_prcp_credit_amount_m6 — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 4.129999876022339]191717931246.47%0.2656
2(4.129999876022339, 33.795000076293945]621059412694.33%0.6892
3(33.795000076293945, 48.94499969482422]443741412966.67%0.2326
4(48.94499969482422, 107.59499740600586]187351707416618.87%-0.0756
5(107.59499740600586, inf]6761596279911.82%-0.3959
2026-11-02T16:27:56.716254 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(6.47% → 11.82% → 11.82%),如需严格单调可考虑调整切点或合并分箱。IV=0.1031,该变量区分度尚可。

WOE 在第 3、4 箱之间穿越 0 线:前 3 箱 WOE>0(客户质量优于整体),后 2 箱 WOE<0(风险高于平均)

第 1→2 箱有局部反弹:Bad Rate 由 6.47% 回落至 4.33%,如需严格单调可考虑合并这两箱

第 5 箱((107.59499740600586, inf])为坏率最高区间:坏率高达 11.82%,可作为拒绝或重点风控策略的阈值参考

cre_now_l30_used_amt_d_min — 3 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 1.0000500082969666]312112827329389.41%-0.0980
2(1.0000500082969666, 1.382200002670288]19331860733.78%0.8757
3(1.382200002670288, inf]27682687812.93%1.1395
2026-11-02T16:27:56.849376 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调下降(9.41% → 2.93%),WOE 单调上升(-0.0980 → 1.1395),符合风控分箱预期。IV=0.1011,该变量区分度尚可。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 2 箱 WOE<0(风险高于平均)

第 1 箱((-inf, 1.0000500082969666])为坏率最高区间:坏率高达 9.41%,可作为拒绝或重点风控策略的阈值参考

loan_os_prcp_credit_amount_m12 — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 16.18000030517578]316430581063.35%0.9564
2(16.18000030517578, 44.21000099182129]603356713626.00%0.3458
3(44.21000099182129, 77.875]144531322312308.51%-0.0308
4(77.875, 95.76499938964844]663560086279.45%-0.1458
5(95.76499938964844, inf]7775695182410.60%-0.2732
2026-11-02T16:27:56.974698 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调上升(3.35% → 10.60%),WOE 单调下降(0.9564 → -0.2732),符合风控分箱预期。IV=0.0894,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 5 箱((95.76499938964844, inf])为坏率最高区间:坏率高达 10.60%,可作为拒绝或重点风控策略的阈值参考

cre_now_l60_used_amt_d_min — 4 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 1.000950038433075]279432520527389.80%-0.1106
2(1.000950038433075, 1.4549500346183777]235622351215.14%0.5858
3(1.4549500346183777, 2.305649995803833]18601789713.82%0.8963
4(2.305649995803833, inf]202619261004.94%0.6276
2026-11-02T16:27:57.099922 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(9.80% → 9.80% → 4.94%),如需严格单调可考虑调整切点或合并分箱。IV=0.0776,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 3→4 箱有局部反弹:Bad Rate 由 3.82% 回升至 4.94%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 1.000950038433075])为坏率最高区间:坏率高达 9.80%,可作为拒绝或重点风控策略的阈值参考

loan_os_prcp_gt0_m3_num_pc — 2 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 68.16500091552734]335432521023.04%1.0563
2(68.16500091552734, inf]347063165930478.78%-0.0649
2026-11-02T16:27:57.213146 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调上升(3.04% → 8.78%),WOE 单调下降(1.0563 → -0.0649),符合风控分箱预期。IV=0.0681,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 1 箱 WOE<0(风险高于平均)

第 2 箱((68.16500091552734, inf])为坏率最高区间:坏率高达 8.78%,可作为拒绝或重点风控策略的阈值参考

cre_now_l90_limit_left_d_min — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, -0.6798500120639801]270025101907.04%0.1917
2(-0.6798500120639801, 0.1424499973654747]288626242629.08%-0.0852
3(0.1424499973654747, 0.9937500059604645]4103356354013.16%-0.5025
4(0.9937500059604645, 1.1872499585151672]791274344786.04%0.3549
5(1.1872499585151672, inf]178551634715088.45%-0.0061
2026-11-02T16:27:57.328926 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(7.04% → 13.16% → 8.45%),如需严格单调可考虑调整切点或合并分箱。IV=0.0635,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 3→4 箱有局部反弹:Bad Rate 由 13.16% 回落至 6.04%,如需严格单调可考虑合并这两箱

第 3 箱((0.1424499973654747, 0.9937500059604645])为坏率最高区间:坏率高达 13.16%,可作为拒绝或重点风控策略的阈值参考

useratemore50maxcontimonsm12 — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]407638891874.59%0.6291
2(0.5, 1.5]200318701336.64%0.2376
3(1.5, 2.5]228621651215.29%0.4787
4(2.5, 4.5]589354014928.35%-0.0099
5(4.5, inf]238022158622169.31%-0.1294
2026-11-02T16:27:57.456180 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(4.59% → 9.31% → 9.31%),如需严格单调可考虑调整切点或合并分箱。IV=0.0578,该变量区分度偏弱但可用。

WOE 在第 3、4 箱之间穿越 0 线:前 3 箱 WOE>0(客户质量优于整体),后 2 箱 WOE<0(风险高于平均)

第 2→3 箱有局部反弹:Bad Rate 由 6.64% 回落至 5.29%,如需严格单调可考虑合并这两箱

第 5 箱((4.5, inf])为坏率最高区间:坏率高达 9.31%,可作为拒绝或重点风控策略的阈值参考

cre_now_p1m_used_rate_m — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, -0.13974999636411667]513748073306.42%0.2755
2(-0.13974999636411667, -4.999999873689376e-05]21512063884.09%0.7513
3(-4.999999873689376e-05, 0.0002500000045984052]1903165724612.93%-0.4959
4(0.0002500000045984052, 0.04154999926686287]429440022926.80%0.2145
5(0.04154999926686287, inf]242492207621738.96%-0.0849
2026-11-02T16:27:57.580020 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(6.42% → 12.93% → 8.96%),如需严格单调可考虑调整切点或合并分箱。IV=0.0577,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 1→2 箱有局部反弹:Bad Rate 由 6.42% 回落至 4.09%,如需严格单调可考虑合并这两箱

第 3 箱((-4.999999873689376e-05, 0.0002500000045984052])为坏率最高区间:坏率高达 12.93%,可作为拒绝或重点风控策略的阈值参考

ps_int_amt_sum — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 209.51499938964844]317330021715.39%0.4594
2(209.51499938964844, 815.2900085449219]1035696177397.14%0.1600
3(815.2900085449219, 1891.7449951171875]1042996138167.82%0.0605
4(1891.7449951171875, 10533.93994140625]113161027610409.19%-0.1154
5(10533.93994140625, inf]2806242238413.69%-0.5642
2026-11-02T16:27:57.705346 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调上升(5.39% → 13.69%),WOE 单调下降(0.4594 → -0.5642),符合风控分箱预期。IV=0.0558,该变量区分度偏弱但可用。

WOE 在第 3、4 箱之间穿越 0 线:前 3 箱 WOE>0(客户质量优于整体),后 2 箱 WOE<0(风险高于平均)

第 5 箱((10533.93994140625, inf])为坏率最高区间:坏率高达 13.69%,可作为拒绝或重点风控策略的阈值参考

cre_now_l60_used_amt_m_max — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 83.30500030517578]514849042444.74%0.5951
2(83.30500030517578, 1001.9599914550781]169371556813698.08%0.0256
3(1001.9599914550781, 2003.010009765625]715165156368.89%-0.0789
4(2003.010009765625, 2696.469970703125]2318205426411.39%-0.3539
5(2696.469970703125, inf]651158746379.78%-0.1840
2026-11-02T16:27:57.831787 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(4.74% → 11.39% → 9.78%),如需严格单调可考虑调整切点或合并分箱。IV=0.0541,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 4→5 箱有局部反弹:Bad Rate 由 11.39% 回落至 9.78%,如需严格单调可考虑合并这两箱

第 4 箱((2003.010009765625, 2696.469970703125])为坏率最高区间:坏率高达 11.39%,可作为拒绝或重点风控策略的阈值参考

cre_l30_l90_used_amt_m_min — 4 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.6299999952316284]298562718526718.95%-0.0853
2(0.6299999952316284, 1107.5999755859375]285827421164.06%0.7573
3(1107.5999755859375, 5082.195068359375]325830821765.40%0.4573
4(5082.195068359375, inf]209319061878.93%-0.0839
2026-11-02T16:27:57.954668 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(8.95% → 8.95% → 8.93%),如需严格单调可考虑调整切点或合并分箱。IV=0.0527,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 2→3 箱有局部反弹:Bad Rate 由 4.06% 回升至 5.40%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 0.6299999952316284])为坏率最高区间:坏率高达 8.95%,可作为拒绝或重点风控策略的阈值参考

avg_tnr 平均履约月数 — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 3.5]24362340963.94%0.7876
2(3.5, 5.5]367234422306.26%0.2998
3(5.5, 6.5]266024721887.07%0.1704
4(6.5, 10.5]126711164510268.10%0.0233
5(10.5, inf]166411503116109.67%-0.1721
2026-11-02T16:27:58.158069 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调上升(3.94% → 9.67%),WOE 单调下降(0.7876 → -0.1721),符合风控分箱预期。IV=0.0524,该变量区分度偏弱但可用。

WOE 在第 4、5 箱之间穿越 0 线:前 4 箱 WOE>0(客户质量优于整体),后 1 箱 WOE<0(风险高于平均)

第 5 箱((10.5, inf])为坏率最高区间:坏率高达 9.67%,可作为拒绝或重点风控策略的阈值参考

cre_now_l30_limit_left_d_min — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.26759999990463257]261323772369.03%-0.0819
2(0.26759999990463257, 0.997050017118454]3523306645712.97%-0.4882
3(0.997050017118454, 1.1242499947547913]137731277010037.28%0.1524
4(1.1242499947547913, 1.7257500290870667]116271058510428.96%-0.0734
5(1.7257500290870667, inf]493946203196.46%0.2813
2026-11-02T16:27:58.286540 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(9.03% → 12.97% → 6.46%),如需严格单调可考虑调整切点或合并分箱。IV=0.0482,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 1→2 箱有局部反弹:Bad Rate 由 9.03% 回升至 12.97%,如需严格单调可考虑合并这两箱

第 2 箱((0.26759999990463257, 0.997050017118454])为坏率最高区间:坏率高达 12.97%,可作为拒绝或重点风控策略的阈值参考

max_loan_os_prcp_m12 — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 1747.2849731445312]410138602415.88%0.3679
2(1747.2849731445312, 2010.25]280625622448.70%-0.0543
3(2010.25, 13572.5849609375]224802076617147.62%0.0888
4(13572.5849609375, 20848.58984375]361932783419.42%-0.1426
5(20848.58984375, inf]5054444560912.05%-0.4180
2026-11-02T16:27:58.434662 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(5.88% → 12.05% → 12.05%),如需严格单调可考虑调整切点或合并分箱。IV=0.0469,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 2→3 箱有局部反弹:Bad Rate 由 8.70% 回落至 7.62%,如需严格单调可考虑合并这两箱

第 5 箱((20848.58984375, inf])为坏率最高区间:坏率高达 12.05%,可作为拒绝或重点风控策略的阈值参考

suc_apprv_num_m6 — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]132721196013129.89%-0.1959
2(0.5, 1.5]873979657748.86%-0.0747
3(1.5, 2.5]477044383326.96%0.1869
4(2.5, 3.5]304228022407.89%0.0515
5(3.5, inf]825777654925.96%0.3530
2026-11-02T16:27:58.568138 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(9.89% → 9.89% → 5.96%),如需严格单调可考虑调整切点或合并分箱。IV=0.0434,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 3→4 箱有局部反弹:Bad Rate 由 6.96% 回升至 7.89%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 0.5])为坏率最高区间:坏率高达 9.89%,可作为拒绝或重点风控策略的阈值参考

useratemore70monthsubcurmonth — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]136451244412018.80%-0.0676
2(0.5, 2.5]658659736139.31%-0.1291
3(2.5, 4.5]519846855139.87%-0.1939
4(4.5, 9.5]603755644737.83%0.0593
5(9.5, inf]659462453495.29%0.4787
2026-11-02T16:27:58.694275 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(8.80% → 9.87% → 5.29%),如需严格单调可考虑调整切点或合并分箱。IV=0.0434,该变量区分度偏弱但可用。

WOE 在第 3、4 箱之间穿越 0 线:前 3 箱 WOE>0(客户质量优于整体),后 2 箱 WOE<0(风险高于平均)

第 1→2 箱有局部反弹:Bad Rate 由 8.80% 回升至 9.31%,如需严格单调可考虑合并这两箱

第 3 箱((2.5, 4.5])为坏率最高区间:坏率高达 9.87%,可作为拒绝或重点风控策略的阈值参考

setl_rem_prcp_paym_m3_num — 2 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]320942924028548.89%-0.0789
2(0.5, inf]596656712954.94%0.5504
2026-11-02T16:27:58.814631 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调下降(8.89% → 4.94%),WOE 单调上升(-0.0789 → 0.5504),符合风控分箱预期。IV=0.0433,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 1 箱 WOE<0(风险高于平均)

第 1 箱((-inf, 0.5])为坏率最高区间:坏率高达 8.89%,可作为拒绝或重点风控策略的阈值参考

orig_prcp_avg_m12 — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 1031.9722290039062]11498107357636.64%0.2381
2(1031.9722290039062, 1871.875]692564055207.51%0.1051
3(1871.875, 3819.857177734375]872779937348.41%-0.0181
4(3819.857177734375, 13787.5]890980418689.74%-0.1798
5(13787.5, inf]2021175626513.11%-0.5149
2026-11-02T16:27:58.956114 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调上升(6.64% → 13.11%),WOE 单调下降(0.2381 → -0.5149),符合风控分箱预期。IV=0.0431,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 5 箱((13787.5, inf])为坏率最高区间:坏率高达 13.11%,可作为拒绝或重点风控策略的阈值参考

useratemore70_m6_num — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]10817101466716.20%0.3103
2(0.5, 1.5]443541153207.22%0.1484
3(1.5, 4.5]144121310413089.08%-0.1013
4(4.5, 5.5]346831323369.69%-0.1734
5(5.5, inf]4928441451410.43%-0.2554
2026-11-02T16:27:59.119782 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调上升(6.20% → 10.43%),WOE 单调下降(0.3103 → -0.2554),符合风控分箱预期。IV=0.0429,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 5 箱((5.5, inf])为坏率最高区间:坏率高达 10.43%,可作为拒绝或重点风控策略的阈值参考

shd_pre_repay_amt_rate — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.6238499879837036]12637118148236.51%0.2584
2(0.6238499879837036, 0.7506499886512756]863179446877.96%0.0421
3(0.7506499886512756, 0.8537499904632568]906682188489.35%-0.1345
4(0.8537499904632568, 0.9255499839782715]5759513162810.90%-0.3052
5(0.9255499839782715, inf]196618031638.29%-0.0022
2026-11-02T16:27:59.250603 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(6.51% → 10.90% → 8.29%),如需严格单调可考虑调整切点或合并分箱。IV=0.0409,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 4→5 箱有局部反弹:Bad Rate 由 10.90% 回落至 8.29%,如需严格单调可考虑合并这两箱

第 4 箱((0.8537499904632568, 0.9255499839782715])为坏率最高区间:坏率高达 10.90%,可作为拒绝或重点风控策略的阈值参考

repay_amount_m6_all_shd_amt_pc — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 93.98500061035156]132571194613119.89%-0.1961
2(93.98500061035156, 95.59499740600586]303328122217.29%0.1378
3(95.59499740600586, 96.48500061035156]297526942819.45%-0.1453
4(96.48500061035156, 99.48500061035156]137351268710487.63%0.0880
5(99.48500061035156, inf]506047722885.69%0.4018
2026-11-02T16:27:59.375349 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(9.89% → 9.89% → 5.69%),如需严格单调可考虑调整切点或合并分箱。IV=0.0386,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 2→3 箱有局部反弹:Bad Rate 由 7.29% 回升至 9.45%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 93.98500061035156])为坏率最高区间:坏率高达 9.89%,可作为拒绝或重点风控策略的阈值参考

cre_now_l30_used_amt_d_max — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.6786499917507172]737668984786.48%0.2678
2(0.6786499917507172, 0.7513499855995178]240421862189.07%-0.0962
3(0.7513499855995178, 0.7632499933242798]191617791377.15%0.1623
4(0.7632499933242798, 0.9139499962329865]174711577117009.73%-0.1740
5(0.9139499962329865, inf]872481106147.04%0.1793
2026-11-02T16:27:59.500021 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(6.48% → 9.73% → 7.04%),如需严格单调可考虑调整切点或合并分箱。IV=0.0362,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 2→3 箱有局部反弹:Bad Rate 由 9.07% 回落至 7.15%,如需严格单调可考虑合并这两箱

第 4 箱((0.7632499933242798, 0.9139499962329865])为坏率最高区间:坏率高达 9.73%,可作为拒绝或重点风控策略的阈值参考

ps_int_amt_avg — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 4.333299875259399]408338572265.54%0.4312
2(4.333299875259399, 13.080849647521973]12227113149137.47%0.1111
3(13.080849647521973, 15.02239990234375]204518412049.98%-0.2060
4(15.02239990234375, 38.24515151977539]123061126510418.46%-0.0244
5(38.24515151977539, inf]7419665376610.32%-0.2443
2026-11-02T16:27:59.689924 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(5.54% → 10.32% → 10.32%),如需严格单调可考虑调整切点或合并分箱。IV=0.0360,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 3→4 箱有局部反弹:Bad Rate 由 9.98% 回落至 8.46%,如需严格单调可考虑合并这两箱

第 5 箱((38.24515151977539, inf])为坏率最高区间:坏率高达 10.32%,可作为拒绝或重点风控策略的阈值参考

useratemore90maxcontimonsm12 — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]929887235756.18%0.3136
2(0.5, 1.5]690363495548.03%0.0332
3(1.5, 4.5]147501346112898.74%-0.0598
4(4.5, 7.5]465241884649.97%-0.2056
5(7.5, inf]2457219026710.87%-0.3013
2026-11-02T16:27:59.816841 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调上升(6.18% → 10.87%),WOE 单调下降(0.3136 → -0.3013),符合风控分箱预期。IV=0.0350,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 5 箱((7.5, inf])为坏率最高区间:坏率高达 10.87%,可作为拒绝或重点风控策略的阈值参考

repay_amount_all_shd_amt_pc — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, -4999.0]215920171426.58%0.2478
2(-4999.0, 90.52499771118164]920783069019.79%-0.1845
3(90.52499771118164, 97.21500015258789]160891468914008.70%-0.0551
4(97.21500015258789, 98.58499908447266]698264825007.16%0.1565
5(98.58499908447266, inf]362334172065.69%0.4029
2026-11-02T16:27:59.939367 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(6.58% → 9.79% → 5.69%),如需严格单调可考虑调整切点或合并分箱。IV=0.0306,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 1→2 箱有局部反弹:Bad Rate 由 6.58% 回升至 9.79%,如需严格单调可考虑合并这两箱

第 2 箱((-4999.0, 90.52499771118164])为坏率最高区间:坏率高达 9.79%,可作为拒绝或重点风控策略的阈值参考

min_apply_amt — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 1135.0]3388301037811.16%-0.3311
2(1135.0, 2270.0]514247344087.93%0.0453
3(2270.0, 4180.0]128411167911629.05%-0.0983
4(4180.0, 7300.0]922184907317.93%0.0463
5(7300.0, inf]748870174716.29%0.2953
2026-11-02T16:28:00.063817 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(11.16% → 11.16% → 6.29%),如需严格单调可考虑调整切点或合并分箱。IV=0.0305,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 2→3 箱有局部反弹:Bad Rate 由 7.93% 回升至 9.05%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 1135.0])为坏率最高区间:坏率高达 11.16%,可作为拒绝或重点风控策略的阈值参考

useratemore90_m3_num — 4 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]232662150117657.59%0.0942
2(0.5, 1.5]611756464717.70%0.0781
3(1.5, 2.5]397936233568.95%-0.0856
4(2.5, inf]4698414155711.86%-0.3996
2026-11-02T16:28:00.180513 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调上升(7.59% → 11.86%),WOE 单调下降(0.0942 → -0.3996),符合风控分箱预期。IV=0.0302,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 2 箱 WOE<0(风险高于平均)

第 4 箱((2.5, inf])为坏率最高区间:坏率高达 11.86%,可作为拒绝或重点风控策略的阈值参考

cre_now_l60_usedrate_m_max — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 4.999999873689376e-05]509447493456.77%0.2166
2(4.999999873689376e-05, 0.039750000461936]263224991335.05%0.5278
3(0.039750000461936, 0.08044999837875366]307928462337.57%0.0971
4(0.08044999837875366, 0.5794999897480011]244692224522249.09%-0.1027
5(0.5794999897480011, inf]279125762157.70%0.0778
2026-11-02T16:28:00.305359 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(6.77% → 9.09% → 7.70%),如需严格单调可考虑调整切点或合并分箱。IV=0.0295,该变量区分度偏弱但可用。

WOE 在第 3、4 箱之间穿越 0 线:前 3 箱 WOE>0(客户质量优于整体),后 2 箱 WOE<0(风险高于平均)

第 1→2 箱有局部反弹:Bad Rate 由 6.77% 回落至 5.05%,如需严格单调可考虑合并这两箱

第 4 箱((0.08044999837875366, 0.5794999897480011])为坏率最高区间:坏率高达 9.09%,可作为拒绝或重点风控策略的阈值参考

min_tnr — 4 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 1.5]268824342549.45%-0.1460
2(1.5, 5.0]202611882614357.08%0.1681
3(5.0, 8.5]301227372759.13%-0.1081
4(8.5, inf]121191093311869.79%-0.1847
2026-11-02T16:28:00.434649 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(9.45% → 9.79% → 9.79%),如需严格单调可考虑调整切点或合并分箱。IV=0.0283,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 1→2 箱有局部反弹:Bad Rate 由 9.45% 回落至 7.08%,如需严格单调可考虑合并这两箱

第 4 箱((8.5, inf])为坏率最高区间:坏率高达 9.79%,可作为拒绝或重点风控策略的阈值参考

loan_os_prcp_gt0_m12_num_pc — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 57.73500061035156]329831381604.85%0.5705
2(57.73500061035156, 72.08000183105469]263124172148.13%0.0186
3(72.08000183105469, 76.01499938964844]196117841779.03%-0.0953
4(76.01499938964844, 92.31500244140625]664060945468.22%0.0067
5(92.31500244140625, inf]235302147820528.72%-0.0575
2026-11-02T16:28:00.558117 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(4.85% → 9.03% → 8.72%),如需严格单调可考虑调整切点或合并分箱。IV=0.0249,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 3→4 箱有局部反弹:Bad Rate 由 9.03% 回落至 8.22%,如需严格单调可考虑合并这两箱

第 3 箱((72.08000183105469, 76.01499938964844])为坏率最高区间:坏率高达 9.03%,可作为拒绝或重点风控策略的阈值参考

grade_limit_amount — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 84500.0]2178194523310.70%-0.2840
2(84500.0, 115500.0]941686567608.07%0.0267
3(115500.0, 137500.0]203319001336.54%0.2533
4(137500.0, 180500.0]6132550163110.29%-0.2406
5(180500.0, inf]183211692813937.60%0.0916
2026-11-02T16:28:00.680125 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(10.70% → 10.70% → 7.60%),如需严格单调可考虑调整切点或合并分箱。IV=0.0226,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 3→4 箱有局部反弹:Bad Rate 由 6.54% 回升至 10.29%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 84500.0])为坏率最高区间:坏率高达 10.70%,可作为拒绝或重点风控策略的阈值参考

end_loan_to_rep_loan_freq — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 1003.5]286772641922587.87%0.0537
2(1003.5, 1935.5]201218191939.59%-0.1626
3(1935.5, 2132.5]2266198428212.44%-0.4550
4(2132.5, 2322.5]266624172499.34%-0.1331
5(2322.5, inf]245922911686.83%0.2068
2026-11-02T16:28:00.802544 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(7.87% → 12.44% → 6.83%),如需严格单调可考虑调整切点或合并分箱。IV=0.0224,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 1→2 箱有局部反弹:Bad Rate 由 7.87% 回升至 9.59%,如需严格单调可考虑合并这两箱

第 3 箱((1935.5, 2132.5])为坏率最高区间:坏率高达 12.44%,可作为拒绝或重点风控策略的阈值参考

shd_pre_repay_cnt — 5 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 25.5]956088766847.15%0.1575
2(25.5, 91.5]11526105489788.49%-0.0275
3(91.5, 237.5]753169775547.36%0.1275
4(237.5, 362.5]253723142238.79%-0.0661
5(362.5, inf]6905619571010.28%-0.2395
2026-11-02T16:28:00.943352 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(7.15% → 10.28% → 10.28%),如需严格单调可考虑调整切点或合并分箱。IV=0.0209,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 2→3 箱有局部反弹:Bad Rate 由 8.49% 回落至 7.36%,如需严格单调可考虑合并这两箱

第 5 箱((362.5, inf])为坏率最高区间:坏率高达 10.28%,可作为拒绝或重点风控策略的阈值参考

last_clear_num_m1 — 3 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, 0.5]301612750926528.79%-0.0667
2(0.5, 1.5]491645943226.55%0.2520
3(1.5, inf]300328271765.86%0.3705
2026-11-02T16:28:01.148948 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调下降(8.79% → 5.86%),WOE 单调上升(-0.0667 → 0.3705),符合风控分箱预期。IV=0.0203,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 2 箱 WOE<0(风险高于平均)

第 1 箱((-inf, 0.5])为坏率最高区间:坏率高达 8.79%,可作为拒绝或重点风控策略的阈值参考

setl_rem_prcp_paym_last_month_lo — 2 箱
分组号分栏TotalGoodBadBad RateWOE
1(-inf, -4999.0]289927571424.90%0.5604
2(-4999.0, inf]351613215430078.55%-0.0361
2026-11-02T16:28:01.265023 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调上升(4.90% → 8.55%),WOE 单调下降(0.5604 → -0.0361),符合风控分箱预期。IV=0.0202,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 1 箱 WOE<0(风险高于平均)

第 2 箱((-4999.0, inf])为坏率最高区间:坏率高达 8.55%,可作为拒绝或重点风控策略的阈值参考

注:统计入模变量经合并调整后的最终分箱明细(开发样本口径);本次无变量发生单调性合并,细/粗分箱一致。

05 变量筛选

项目内容
筛选优先级黑名单 > 白名单 > 自动规则
IV 下限0.02
缺失率上限0.8
集中度上限0.9
相关性上限(皮尔逊)0.7
PSI 上限0.25
重要性 Top-K80
黑名单无
白名单无

候选变量 124,成功分箱 124,入模 47

变量白名单入模IV缺失率集中度IV门槛相关性PSI剔除原因
ps_perd_no—N0.1475PASSPASSPASS—0.3217PSI>0.25
age—N0.0051PASSPASSFAIL——IV<0.02
max_int_rat—N0.0172PASSPASSFAIL——IV<0.02
min_int_rat—N0.0050PASSPASSFAIL——IV<0.02
avg_tnr
平均履约月数
—Y0.0524PASSPASSPASSPASS0.0410
min_tnr—Y0.0283PASSPASSPASSPASS0.0040
sum_apply_amt—N0.0142PASSPASSFAIL——IV<0.02
min_apply_amt—Y0.0305PASSPASSPASSPASS0.0271
min_apprv_amt—N0.0124PASSPASSFAIL——IV<0.02
apprv_amt_day—Y0.1569PASSPASSPASSPASS0.1356
end_loan_to_rep_loan_freq—Y0.0224PASSPASSPASSPASS0.0395
end_loan_to_cur_freq—N0.1904PASSPASSPASS—0.2531PSI>0.25
f_rep_loan_to_f_apply_dt_freq—N0.0054PASSPASSFAIL——IV<0.02
ps_prcp_amt_mon—N0.0282PASSPASSPASS—0.3356PSI>0.25
ps_int_amt_mon—N0.0946PASSPASSPASS—0.3037PSI>0.25
ps_int_amt_sum—Y0.0558PASSPASSPASSPASS0.2234
ps_prcp_amt_avg—N0.0094PASSPASSFAIL——IV<0.02
ps_int_amt_avg—Y0.0360PASSPASSPASSPASS0.1191
setl_prcp_mon_ratio—N0.0149PASSPASSFAIL——IV<0.02
acc_ps_perd_no—N0.0987PASSPASSPASS—0.4588PSI>0.25
max_od_days—N0.2818PASSPASSPASS—0.5635PSI>0.25
m_s_ps_instm_amt—N0.0386PASSPASSPASS—0.2760PSI>0.25
f_ps_od_days_max—Y0.1937PASSPASSPASSPASS0.1507
s_ps_od_days_max—Y0.1336PASSPASSPASSPASS0.1952
t_ps_od_days_max—Y0.1496PASSPASSPASSPASS0.2430
ovd_days_d3_num_m1—N0.1789PASSFAIL———集中度>0.9
orig_prcp_avg_m3—N0.1453PASSPASSPASS—0.2678PSI>0.25
orig_prcp_avg_m12—Y0.0431PASSPASSPASSPASS0.0373
suc_apply_trench_num—N0.0091PASSFAIL———集中度>0.9
suc_apply_top_pc_code_num—N0.0022PASSPASSFAIL——IV<0.02
suc_apprv_num_m1—Y0.1594PASSPASSPASSPASS0.1386
apprv_num_m1—Y0.1225PASSPASSPASSPASS0.1088
suc_apply_num_m3—N0.0000PASSFAIL———集中度>0.9
apprv_num_m3—N0.0970PASSPASSPASSFAIL0.2416与 apprv_num_m1 相关>0.7
suc_apprv_num_m6—Y0.0434PASSPASSPASSPASS0.1217
apply_num_m12—N0.0007PASSPASSFAIL——IV<0.02
last_apply_od_num_m12—N0.0067PASSFAIL———集中度>0.9
last_apply_freq—N0.0173PASSPASSFAIL——IV<0.02
dec_loan_num—N0.0028PASSFAIL———集中度>0.9
bh_cur_od_num—N0.0064PASSPASSFAIL——IV<0.02
hj_cur_od_num—N0.0005PASSFAIL———集中度>0.9
setl_prcp_ratio—N0.1275PASSPASSPASS—0.2983PSI>0.25
last_loan_setl_tnr—N0.1645PASSPASSPASS—0.2903PSI>0.25
last_od_num_m6—N0.1099PASSPASSPASS—0.6013PSI>0.25
last_od_to_cur_freq—N0.2596PASSPASSPASS—0.8376PSI>0.25
last_clear_num_m1—Y0.0203PASSPASSPASSPASS0.0046
last_clear_num_m3—N0.0305PASSPASSPASSFAIL0.0139与 suc_apprv_num_m6 相关>0.7
grade_limit_amount—Y0.0226PASSPASSPASSPASS0.2077
exposure—N0.0155PASSPASSFAIL——IV<0.02
available_limit_amt—Y0.1061PASSPASSPASSPASS0.0138
account_age—N0.0164PASSPASSFAIL——IV<0.02
account_age_loan_day—N0.0134PASSPASSFAIL——IV<0.02
max_od_due—N0.0638PASSPASSPASS—0.4345PSI>0.25
odgt0_m12_mob_pc—N0.0450PASSPASSPASS—0.9610PSI>0.25
odgt0_m3_num—N0.0227PASSPASSPASS—2.4218PSI>0.25
odgt0_m12_num—N0.0464PASSPASSPASS—0.9376PSI>0.25
odgt0_m12_subcurmonth—N0.2143PASSPASSPASS—2.5210PSI>0.25
tothismonthodcontiaddnumm12—N0.0187PASSPASSFAIL——IV<0.02
tothismonthodcontireducenumm12—N0.0209PASSPASSPASS—0.7206PSI>0.25
od_amt_m6_loan_os_prcp_m6_pc—N0.0804PASSFAIL———集中度>0.9
od_amt_m12_loan_os_prcp_m12_pc—N0.0658PASSFAIL———集中度>0.9
avg_od_amt_m6_orig_prcp_pc—N0.0815PASSFAIL———集中度>0.9
max_loan_os_prcp_m12—Y0.0469PASSPASSPASSPASS0.0638
loan_os_prcp_gt0_m3_num_pc—Y0.0681PASSPASSPASSPASS0.0255
loan_os_prcp_gt0_m12_num_pc—Y0.0249PASSPASSPASSPASS0.0370
repay_amount_m3_loan_os_prcp_pc—N0.0661PASSPASSPASS—0.3259PSI>0.25
repay_amount_m6_loan_os_prcp_pc—N0.0294PASSPASSPASS—0.2694PSI>0.25
repay_amount_all_shd_amt_pc—Y0.0306PASSPASSPASSPASS0.2346
repay_amount_m6_all_shd_amt_pc—Y0.0386PASSPASSPASSPASS0.0409
setl_rem_prcp_paym_last_month_lo—Y0.0202PASSPASSPASSPASS0.1082
setl_rem_prcp_paym_m12_loan_os_p—N0.0086PASSPASSFAIL——IV<0.02
setl_rem_prcp_paym_m3_num—Y0.0433PASSPASSPASSPASS0.0163
setl_rem_prcp_paym_m12_num—N0.0049PASSPASSFAIL——IV<0.02
loan_os_prcp_credit_amount_m3—Y0.1276PASSPASSPASSPASS0.0733
loan_os_prcp_credit_amount_m6—Y0.1031PASSPASSPASSPASS0.0433
loan_os_prcp_credit_amount_m12—Y0.0894PASSPASSPASSPASS0.0786
useratemore50maxcontimonsm12—Y0.0578PASSPASSPASSPASS0.2289
useratemore70_m6_num—Y0.0429PASSPASSPASSPASS0.0784
useratemore70monthsubcurmonth—Y0.0434PASSPASSPASSPASS0.1603
useratemore90_m3_num—Y0.0302PASSPASSPASSPASS0.0027
useratemore90maxcontimonsm12—Y0.0350PASSPASSPASSPASS0.1030
shd_max_dpd—N0.0690PASSPASSPASS—0.3133PSI>0.25
shd_min_dpd—N0.0130PASSPASSFAIL——IV<0.02
shd_pre_repay_cnt—Y0.0209PASSPASSPASSPASS0.0547
shd_perd_max—N0.0161PASSFAIL———集中度>0.9
shd_instm_amt_sum—N0.0298PASSPASS———唯一率过高(ID列)
shd_l30_max_dpd—N0.0576PASSPASSPASS—1.1956PSI>0.25
shd_l60_min_dpd—N0.0368PASSPASSPASS—0.5202PSI>0.25
shd_l90_ps_due_cnt—N0.0336PASSPASSPASS—0.3409PSI>0.25
shd_l180_max_dpd—N0.0043PASSPASSFAIL——IV<0.02
shd_l180_instm_amt_sum—N0.0243PASSPASS———唯一率过高(ID列)
shd_remain_actual_cnt—N0.0182PASSPASSFAIL——IV<0.02
shd_pre_repay_cnt_rate—N0.0124PASSPASSFAIL——IV<0.02
shd_pre_repay_amt_rate—Y0.0409PASSPASSPASSPASS0.0403
shd_pre_repay_30d_amt_rate—N0.0062PASSPASSFAIL——IV<0.02
shd_pre_repay_due_rate—N0.1142PASSPASSPASS—0.3521PSI>0.25
shd_l30_pre_repay_due_rate—N0.0546PASSPASSPASS—0.8956PSI>0.25
shd_remain_actual_rate—N0.0110PASSPASSFAIL——IV<0.02
cre_usedrate_u30_daydelta_min—N0.0134PASSPASSFAIL——IV<0.02
cre_usedrate_u30_daydelta_max—N0.0468PASSPASSPASS—0.2756PSI>0.25
cre_l30_limit_left_amt_min—N0.0597PASSPASSPASSFAIL0.0141与 available_limit_amt 相关>0.7
cre_l60_used_limit_rate_max—N0.0767PASSPASSPASSFAIL0.0654与 cre_l60_used_limit_rate_min 相关>0.7
cre_l60_used_limit_rate_min—Y0.1478PASSPASSPASSPASS0.0410
cre_l90_used_limit_amt_min—Y0.1336PASSPASSPASSPASS0.0680
cre_now_p1m_used_rate_d—N0.0563PASSPASSPASSFAIL0.0765与 cre_now_l30_used_amt_d_min 相关>0.7
cre_now_p3m_used_rate_d—N0.0274PASSPASSPASSFAIL0.0513与 cre_now_l60_used_amt_d_min 相关>0.7
cre_now_p1m_used_rate_m—Y0.0577PASSPASSPASSPASS0.0782
cre_now_p2m_used_rate_m—N0.0265PASSPASSPASSFAIL0.0446与 cre_now_p1m_used_rate_m 相关>0.7
cre_now_l30_used_amt_d_max—Y0.0362PASSPASSPASSPASS0.2379
cre_now_l30_used_amt_d_min—Y0.1011PASSPASSPASSPASS0.0345
cre_now_l30_limit_left_d_max—N0.1167PASSPASSPASSFAIL0.0180与 cre_now_l60_limit_left_d_max 相关>0.7
cre_now_l30_limit_left_d_min—Y0.0482PASSPASSPASSPASS0.0922
cre_now_l60_used_amt_d_min—Y0.0776PASSPASSPASSPASS0.0609
cre_now_l60_limit_left_d_max—Y0.1208PASSPASSPASSPASS0.0217
cre_now_l60_limit_left_d_min—N0.0600PASSPASSPASSFAIL0.1481与 cre_now_l90_limit_left_d_min 相关>0.7
cre_now_l60_used_amt_m_max—Y0.0541PASSPASSPASSPASS0.2428
cre_now_l60_used_amt_m_min—Y0.1485PASSPASSPASSPASS0.0815
cre_now_l60_usedrate_m_max—Y0.0295PASSPASSPASSPASS0.2041
cre_now_l90_limit_left_d_min—Y0.0635PASSPASSPASSPASS0.1869
cre_l30_l90_used_amt_d_max—N0.0993PASSPASSPASS—0.3521PSI>0.25
cre_l30_l90_used_amt_m_max—N0.0882PASSPASSPASS—0.3031PSI>0.25
cre_l30_l90_used_amt_m_min—Y0.0527PASSPASSPASSPASS0.0608
cre_l30_l90_usedrate_m_max—N0.0258PASSPASSPASSFAIL0.2046与 cre_now_l60_usedrate_m_max 相关>0.7
cre_l30_l90_usedrate_m_min—N0.0119PASSPASSFAIL——IV<0.02

06 回归结果

待补充:本实验未运行 LR 路线(或未产生回归结果),无回归明细。

07 VIF

待补充:本实验未运行 LR 路线,未计算 VIF/相关系数。

08 最终评分卡

待补充:本实验未生成评分卡(未运行 LR 路线)。

09 模型表现

模型区分度评估(按样本)

模型样本KSAUCGINI
LightGBM开发样本0.33090.72240.4448
LightGBM测试样本0.30930.71360.4272
LightGBM训练+测试样本0.32200.71880.4376
LightGBM跨时间验证样本0.30400.70820.4164

LightGBM — 4 样本评估表(score_cut 粒度)

开发样本(KS 0.3309,AUC 0.7224)

cutscntbadgoodbad_ratepopulation_pctcum_bad_ratecum_good_rate
[-inf, 552)3809844296522.16%10.00%26.79%8.49%
[552, 559)3809549326014.41%10.00%44.21%17.82%
[559, 564)3809460334912.08%10.00%58.81%27.40%
[564, 568)380834734619.11%10.00%69.82%37.31%
[568, 573)380928135287.38%10.00%78.74%47.41%
[573, 578)380921135985.54%10.00%85.43%57.71%
[578, 584)380818536234.86%10.00%91.30%68.08%
[584, 591)380915336564.02%10.00%96.16%78.54%
[591, 598)38099037192.36%10.00%99.02%89.19%
[598, inf)38093137780.81%10.00%100.00%100.00%
2026-11-02T16:28:01.398506 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/ 2026-11-02T16:28:01.529808 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

测试样本(KS 0.3093,AUC 0.7136)

cutscntbadgoodbad_ratepopulation_pctcum_bad_ratecum_good_rate
[-inf, 552)2509510199920.33%9.88%24.29%8.58%
[552, 559)2569388218115.10%10.12%42.76%17.95%
[559, 564)2506311219512.41%9.87%57.57%27.37%
[564, 568)262723023978.76%10.35%68.52%37.66%
[568, 573)255019923517.80%10.04%78.00%47.76%
[573, 578)243415822766.49%9.59%85.52%57.53%
[578, 584)246312323404.99%9.70%91.38%67.58%
[584, 591)265210225503.85%10.44%96.24%78.52%
[591, 598)25946125332.35%10.22%99.14%89.40%
[598, inf)24871824690.72%9.79%100.00%100.00%
2026-11-02T16:28:01.670983 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/ 2026-11-02T16:28:01.806666 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

训练+测试样本(KS 0.3220,AUC 0.7188)

cutscntbadgoodbad_ratepopulation_pctcum_bad_ratecum_good_rate
[-inf, 552)63181354496421.43%9.95%25.79%8.53%
[552, 559)6378937544114.69%10.05%43.63%17.87%
[559, 564)6315771554412.21%9.95%58.31%27.39%
[564, 568)643557758588.97%10.14%69.30%37.45%
[568, 573)635948058797.55%10.02%78.44%47.55%
[573, 578)624336958745.91%9.83%85.47%57.64%
[578, 584)627130859634.91%9.88%91.33%67.88%
[584, 591)646125562063.95%10.18%96.19%78.53%
[591, 598)640315162522.36%10.09%99.07%89.27%
[598, inf)62964962470.78%9.92%100.00%100.00%
2026-11-02T16:28:01.939648 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/ 2026-11-02T16:28:02.070207 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

跨时间验证样本(KS 0.3040,AUC 0.7082)

cutscntbadgoodbad_ratepopulation_pctcum_bad_ratecum_good_rate
[-inf, 552)3445652279318.93%9.92%26.81%8.65%
[552, 559)3392420297212.38%9.77%44.08%17.85%
[559, 564)351432531899.25%10.12%57.44%27.72%
[564, 568)355525133047.06%10.24%67.76%37.95%
[568, 573)323118830435.82%9.30%75.49%47.37%
[573, 578)353419733375.57%10.18%83.59%57.71%
[578, 584)363316534684.54%10.46%90.38%68.44%
[584, 591)341711133063.25%9.84%94.94%78.68%
[591, 598)32987532232.27%9.50%98.03%88.66%
[598, inf)37114836631.29%10.69%100.00%100.00%
2026-11-02T16:28:02.207477 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/ 2026-11-02T16:28:02.340733 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

跨时间验证样本结论

风险高度集中在低分尾端——最低 20% 样本集中了全量 57.4% 的坏客户,策略切分空间很大。

① 样本与分箱结构:共 34,730 样本 / 2,432 坏样本,整体坏率 7.00%,好坏比约 13.3:1;10 档等频分数分栏,每箱约 10%,分箱均匀,无空箱无过厚箱。

② 排序性与单调性:逐箱 Bad Rate 18.93% → … → 1.29%(低分 → 高分),无倒挂,最低分段坏率是最高分段的 14.7 倍。

③ 区分度与坏客户捕获:最低 20% 样本捕获全量 57.4% 的坏客户,最低 50% 捕获 83.6%;最高 20% 样本仅含 2.0% 的坏客户,其中风险最高一箱单独占 26.8%。

(口径:分数 = 按基准分/优比/PDO 刻度由模型概率转换,分数越高风险越低;结论由程序按分箱数据自动生成。)

注:score_cut 表按开发样本(train)分数等频 10 档取边界(低分 → 高分),各样本共用同一套边界、同栏可比(20261022 口径);分数由模型概率按基准分/优比/PDO 刻度转换,分数越高风险越低;每表附 KS/AUC(按概率计算,与分栏口径无关);训练+测试 = TRAIN 全体。旧实验无分数转换时为 PD 等频口径(按本样本分位切分)。

10 KS和GINI

LightGBM

样本KSGINIAUC
开发样本0.33090.44480.7224
测试样本0.30930.42720.7136
训练+测试样本0.32200.43760.7188
跨时间验证样本0.30400.41640.7082

开发样本

2026-11-02T16:28:02.507276 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/ 2026-11-02T16:28:02.580972 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

测试样本

2026-11-02T16:28:02.654916 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/ 2026-11-02T16:28:02.730632 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

训练+测试样本

2026-11-02T16:28:02.803674 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/ 2026-11-02T16:28:02.878987 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

跨时间验证样本

2026-11-02T16:28:02.952107 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/ 2026-11-02T16:28:03.027975 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

注:GINI = 2×AUC−1。KS 图 X 轴为分数(升序),红虚线为最大间隔位置;ROC 图例标注 AUC。

11 PSI

模型稳定性评估

模型训练集与测试集训练集与跨时间验证训练集与oot验证集训练集与psi验证集
LightGBM0.00070.0016—0.0089

分数段 PSI 明细(开发样本 vs 跨时间验证样本)

分数段开发样本占比跨时间验证样本占比PSI 贡献累计 PSI
[-inf, 552)10.00%9.92%0.00000.0000
[552, 559)10.00%9.77%0.00010.0001
[559, 564)10.00%10.12%0.00000.0001
[564, 568)10.00%10.24%0.00010.0001
[568, 573)10.00%9.30%0.00050.0006
[573, 578)10.00%10.18%0.00000.0007
[578, 584)10.00%10.46%0.00020.0009
[584, 591)10.00%9.84%0.00000.0009
[591, 598)10.00%9.50%0.00030.0012
[598, inf)10.00%10.69%0.00050.0016

分数段 PSI 明细(开发样本 vs PSI时点样本)

分数段开发样本占比PSI时点占比PSI 贡献累计 PSI
[-inf, 552)10.00%10.78%0.00060.0006
[552, 559)10.00%10.18%0.00000.0006
[559, 564)10.00%10.58%0.00030.0009
[564, 568)10.00%10.03%0.00000.0009
[568, 573)10.00%9.89%0.00000.0010
[573, 578)10.00%10.38%0.00010.0011
[578, 584)10.00%11.03%0.00100.0021
[584, 591)10.00%10.56%0.00030.0024
[591, 598)10.00%8.22%0.00350.0059
[598, inf)10.00%8.35%0.00300.0089
2026-11-02T16:28:03.116198 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

注:分数段 PSI 明细按分数刻度计算(概率→分数转换,20261022 口径),分栏边界取自开发样本(train)分数等频分位,各对比样本共用同一套边界。

模型样本量分布评估

评分段开发样本样本量开发样本占比测试样本样本量测试样本占比训练+测试样本样本量训练+测试样本占比跨时间验证样本样本量跨时间验证样本占比PSI时点样本样本量PSI时点样本占比
(-inf, 552]380910.00%25099.88%63189.95%34459.92%188310.78%
(552, 559]380910.00%256910.12%637810.05%33929.77%177810.18%
(559, 564]380910.00%25069.87%63159.95%351410.12%184910.58%
(564, 568]380810.00%262710.35%643510.14%355510.24%175310.03%
(568, 573]380910.00%255010.04%635910.02%32319.30%17289.89%
(573, 578]380910.00%24349.59%62439.83%353410.18%181410.38%
(578, 584]380810.00%24639.70%62719.88%363310.46%192711.03%
(584, 591]380910.00%265210.44%646110.18%34179.84%184610.56%
(591, 598]380910.00%259410.22%640310.09%32989.50%14368.22%
(598, inf]380910.00%24879.79%62969.92%371110.69%14598.35%
sum38088100.00%25391100.00%63479100.00%34730100.00%17473100.00%

变量 PSI(开发样本 vs PSI时点样本,按入模变量分箱)

特征项PSI状态
t_ps_od_days_max0.2430关注
cre_now_l60_used_amt_m_max0.2428关注
cre_now_l30_used_amt_d_max0.2379关注
repay_amount_all_shd_amt_pc0.2346关注
useratemore50maxcontimonsm120.2289关注
ps_int_amt_sum0.2234关注
grade_limit_amount0.2077关注
cre_now_l60_usedrate_m_max0.2041关注
s_ps_od_days_max0.1952关注
cre_now_l90_limit_left_d_min0.1869关注
useratemore70monthsubcurmonth0.1603关注
f_ps_od_days_max0.1507关注
suc_apprv_num_m10.1386关注
apprv_amt_day0.1356关注
suc_apprv_num_m60.1217关注
ps_int_amt_avg0.1191关注
apprv_num_m10.1088关注
setl_rem_prcp_paym_last_month_lo0.1082关注
useratemore90maxcontimonsm120.1030关注
cre_now_l30_limit_left_d_min0.0922稳定
cre_now_l60_used_amt_m_min0.0815稳定
loan_os_prcp_credit_amount_m120.0786稳定
useratemore70_m6_num0.0784稳定
cre_now_p1m_used_rate_m0.0782稳定
loan_os_prcp_credit_amount_m30.0733稳定
cre_l90_used_limit_amt_min0.0680稳定
max_loan_os_prcp_m120.0638稳定
cre_now_l60_used_amt_d_min0.0609稳定
cre_l30_l90_used_amt_m_min0.0608稳定
shd_pre_repay_cnt0.0547稳定
loan_os_prcp_credit_amount_m60.0433稳定
avg_tnr
平均履约月数
0.0410稳定
cre_l60_used_limit_rate_min0.0410稳定
repay_amount_m6_all_shd_amt_pc0.0409稳定
shd_pre_repay_amt_rate0.0403稳定
end_loan_to_rep_loan_freq0.0395稳定
orig_prcp_avg_m120.0373稳定
loan_os_prcp_gt0_m12_num_pc0.0370稳定
cre_now_l30_used_amt_d_min0.0345稳定
min_apply_amt0.0271稳定
loan_os_prcp_gt0_m3_num_pc0.0255稳定
cre_now_l60_limit_left_d_max0.0217稳定
setl_rem_prcp_paym_m3_num0.0163稳定
available_limit_amt0.0138稳定
last_clear_num_m10.0046稳定
min_tnr0.0040稳定
useratemore90_m3_num0.0027稳定
f_ps_od_days_max — PSI 0.1507
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 0.5]53.63%71.67%18.04%0.0523
(0.5, 1.5]22.17%15.68%-6.49%0.0225
(1.5, 2.5]8.98%5.01%-3.97%0.0232
(2.5, 4.5]9.43%4.97%-4.45%0.0285
(4.5, inf]5.80%2.67%-3.12%0.0242
suc_apprv_num_m1 — PSI 0.1386
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 0.5]79.23%62.50%-16.73%0.0397
(0.5, 1.5]12.57%22.98%10.41%0.0628
(1.5, inf]8.20%14.52%6.32%0.0361
apprv_amt_day — PSI 0.1356
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 11.0]79.68%63.48%-16.21%0.0369
(11.0, 1000.5]6.58%11.57%4.99%0.0281
(1000.5, 5995.0]8.68%14.12%5.44%0.0265
(5995.0, inf]5.06%10.84%5.78%0.0441
t_ps_od_days_max — PSI 0.2430
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 0.5]54.18%75.91%21.72%0.0732
(0.5, 2.5]29.14%17.99%-11.15%0.0538
(2.5, 3.5]5.38%2.30%-3.08%0.0261
(3.5, 5.5]5.77%2.56%-3.21%0.0261
(5.5, inf]5.53%1.25%-4.28%0.0637
cre_now_l60_used_amt_m_min — PSI 0.0815
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 1.7799999713897705]74.09%61.70%-12.38%0.0227
(1.7799999713897705, 869.5950012207031]6.50%8.10%1.59%0.0035
(869.5950012207031, 1999.7100219726562]5.41%7.26%1.84%0.0054
(1999.7100219726562, 6462.93505859375]8.21%11.44%3.23%0.0107
(6462.93505859375, inf]5.79%11.50%5.72%0.0393
cre_l60_used_limit_rate_min — PSI 0.0410
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 0.019700000062584877]11.67%16.73%5.06%0.0182
(0.019700000062584877, 0.09174999967217445]7.39%8.24%0.84%0.0009
(0.09174999967217445, 0.2787500023841858]19.24%18.81%-0.43%0.0001
(0.2787500023841858, 0.9161500036716461]49.65%41.27%-8.38%0.0155
(0.9161500036716461, inf]12.06%14.96%2.90%0.0063
s_ps_od_days_max — PSI 0.1952
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 0.5]54.12%74.07%19.95%0.0626
(0.5, 1.5]20.88%14.30%-6.58%0.0249
(1.5, 2.5]9.10%4.84%-4.26%0.0269
(2.5, 4.5]9.18%4.50%-4.68%0.0333
(4.5, inf]6.72%2.30%-4.42%0.0474
cre_l90_used_limit_amt_min — PSI 0.0680
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 21.640000343322754]14.76%21.47%6.71%0.0251
(21.640000343322754, 748.7099914550781]21.22%17.21%-4.01%0.0084
(748.7099914550781, 8516.22998046875]52.99%45.70%-7.30%0.0108
(8516.22998046875, 14597.384765625]5.96%6.57%0.61%0.0006
(14597.384765625, inf]5.08%9.06%3.99%0.0231
loan_os_prcp_credit_amount_m3 — PSI 0.0733
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, -4999.0]9.66%6.42%-3.24%0.0133
(-4999.0, 35.125]21.85%26.69%4.85%0.0097
(35.125, 100.47500228881836]50.23%48.56%-1.67%0.0006
(100.47500228881836, 124.34499740600586]7.14%2.99%-4.15%0.0362
(124.34499740600586, inf]11.13%15.35%4.22%0.0136
apprv_num_m1 — PSI 0.1088
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 0.5]73.93%58.74%-15.19%0.0349
(0.5, 1.5]14.05%24.06%10.01%0.0539
(1.5, 2.5]5.22%8.35%3.13%0.0147
(2.5, inf]6.80%8.85%2.05%0.0054
cre_now_l60_limit_left_d_max — PSI 0.0217
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, -0.013299999758601189]5.94%6.41%0.46%0.0003
(-0.013299999758601189, 0.3223000019788742]10.38%11.03%0.65%0.0004
(0.3223000019788742, 0.6971500217914581]6.32%7.61%1.28%0.0024
(0.6971500217914581, 0.9999499917030334]7.62%11.15%3.52%0.0134
(0.9999499917030334, inf]69.73%63.81%-5.92%0.0053
available_limit_amt — PSI 0.0138
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 0.03499999921768904]32.52%32.39%-0.14%0.0000
(0.03499999921768904, 468.864990234375]11.88%9.29%-2.59%0.0064
(468.864990234375, 4307.300048828125]31.51%30.87%-0.64%0.0001
(4307.300048828125, 27316.265625]19.04%20.46%1.42%0.0010
(27316.265625, inf]5.05%6.99%1.94%0.0063
loan_os_prcp_credit_amount_m6 — PSI 0.0433
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 4.129999876022339]5.04%6.36%1.32%0.0031
(4.129999876022339, 33.795000076293945]16.32%23.19%6.88%0.0242
(33.795000076293945, 48.94499969482422]11.66%12.11%0.45%0.0002
(48.94499969482422, 107.59499740600586]49.22%40.80%-8.43%0.0158
(107.59499740600586, inf]17.76%17.55%-0.21%0.0000
cre_now_l30_used_amt_d_min — PSI 0.0345
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 1.0000500082969666]86.91%80.18%-6.73%0.0054
(1.0000500082969666, 1.382200002670288]5.38%7.35%1.96%0.0061
(1.382200002670288, inf]7.71%12.47%4.76%0.0229
loan_os_prcp_credit_amount_m12 — PSI 0.0786
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 16.18000030517578]8.31%12.94%4.63%0.0205
(16.18000030517578, 44.21000099182129]15.85%21.83%5.98%0.0192
(44.21000099182129, 77.875]37.97%32.44%-5.53%0.0087
(77.875, 95.76499938964844]17.43%11.02%-6.41%0.0294
(95.76499938964844, inf]20.43%21.76%1.33%0.0008
cre_now_l60_used_amt_d_min — PSI 0.0609
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 1.000950038433075]81.74%71.70%-10.04%0.0132
(1.000950038433075, 1.4549500346183777]6.89%9.65%2.76%0.0093
(1.4549500346183777, 2.305649995803833]5.44%7.75%2.31%0.0081
(2.305649995803833, inf]5.93%10.90%4.98%0.0303
loan_os_prcp_gt0_m3_num_pc — PSI 0.0255
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 68.16500091552734]8.81%13.83%5.02%0.0226
(68.16500091552734, inf]91.19%86.17%-5.02%0.0028
cre_now_l90_limit_left_d_min — PSI 0.1869
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, -0.6798500120639801]7.62%4.88%-2.73%0.0121
(-0.6798500120639801, 0.1424499973654747]8.14%4.66%-3.48%0.0194
(0.1424499973654747, 0.9937500059604645]11.57%12.02%0.44%0.0002
(0.9937500059604645, 1.1872499585151672]22.31%41.19%18.88%0.1157
(1.1872499585151672, inf]50.36%37.25%-13.11%0.0395
useratemore50maxcontimonsm12 — PSI 0.2289
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 0.5]10.71%20.26%9.55%0.0609
(0.5, 1.5]5.26%11.54%6.27%0.0492
(1.5, 2.5]6.01%10.12%4.11%0.0215
(2.5, 4.5]15.48%17.49%2.00%0.0024
(4.5, inf]62.54%40.59%-21.94%0.0948
cre_now_p1m_used_rate_m — PSI 0.0782
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, -0.13974999636411667]13.61%15.70%2.09%0.0030
(-0.13974999636411667, -4.999999873689376e-05]5.70%7.59%1.89%0.0054
(-4.999999873689376e-05, 0.0002500000045984052]5.04%11.18%6.13%0.0488
(0.0002500000045984052, 0.04154999926686287]11.38%12.22%0.84%0.0006
(0.04154999926686287, inf]64.26%53.31%-10.95%0.0205
ps_int_amt_sum — PSI 0.2234
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 209.51499938964844]8.33%24.84%16.51%0.1804
(209.51499938964844, 815.2900085449219]27.20%19.86%-7.33%0.0230
(815.2900085449219, 1891.7449951171875]27.39%20.94%-6.45%0.0173
(1891.7449951171875, 10533.93994140625]29.72%26.96%-2.75%0.0027
(10533.93994140625, inf]7.37%7.39%0.03%0.0000
cre_now_l60_used_amt_m_max — PSI 0.2428
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 83.30500030517578]13.52%32.92%19.40%0.1726
(83.30500030517578, 1001.9599914550781]44.49%31.12%-13.38%0.0479
(1001.9599914550781, 2003.010009765625]18.79%13.42%-5.37%0.0181
(2003.010009765625, 2696.469970703125]6.09%4.64%-1.45%0.0040
(2696.469970703125, inf]17.11%17.90%0.80%0.0004
cre_l30_l90_used_amt_m_min — PSI 0.0608
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 0.6299999952316284]78.43%68.57%-9.86%0.0132
(0.6299999952316284, 1107.5999755859375]7.51%9.46%1.95%0.0045
(1107.5999755859375, 5082.195068359375]8.56%11.12%2.56%0.0067
(5082.195068359375, inf]5.50%10.84%5.34%0.0363
avg_tnr — PSI 0.0410
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 3.5]6.40%11.49%5.10%0.0298
(3.5, 5.5]9.64%9.21%-0.43%0.0002
(5.5, 6.5]6.99%8.13%1.15%0.0017
(6.5, 10.5]33.27%27.97%-5.30%0.0092
(10.5, inf]43.70%43.19%-0.51%0.0001
cre_now_l30_limit_left_d_min — PSI 0.0922
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 0.26759999990463257]7.16%3.86%-3.30%0.0204
(0.26759999990463257, 0.997050017118454]9.66%11.11%1.45%0.0020
(0.997050017118454, 1.1242499947547913]37.76%50.32%12.56%0.0361
(1.1242499947547913, 1.7257500290870667]31.88%22.55%-9.32%0.0323
(1.7257500290870667, inf]13.54%12.16%-1.38%0.0015
max_loan_os_prcp_m12 — PSI 0.0638
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 1747.2849731445312]10.78%16.33%5.56%0.0231
(1747.2849731445312, 2010.25]7.37%5.85%-1.52%0.0035
(2010.25, 13572.5849609375]59.06%48.73%-10.34%0.0199
(13572.5849609375, 20848.58984375]9.51%10.92%1.41%0.0020
(20848.58984375, inf]13.28%18.17%4.89%0.0154
suc_apprv_num_m6 — PSI 0.1217
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 0.5]34.85%20.09%-14.76%0.0813
(0.5, 1.5]22.95%32.15%9.20%0.0310
(1.5, 2.5]12.53%15.76%3.23%0.0074
(2.5, 3.5]7.99%8.25%0.26%0.0001
(3.5, inf]21.68%23.76%2.07%0.0019
useratemore70monthsubcurmonth — PSI 0.1603
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 0.5]35.85%41.31%5.46%0.0077
(0.5, 2.5]17.30%11.47%-5.84%0.0240
(2.5, 4.5]13.66%6.52%-7.14%0.0527
(4.5, 9.5]15.86%11.51%-4.35%0.0139
(9.5, inf]17.33%29.18%11.86%0.0618
setl_rem_prcp_paym_m3_num — PSI 0.0163
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 0.5]84.32%79.42%-4.90%0.0029
(0.5, inf]15.68%20.58%4.90%0.0133
orig_prcp_avg_m12 — PSI 0.0373
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 1031.9722290039062]30.19%28.38%-1.82%0.0011
(1031.9722290039062, 1871.875]18.19%14.83%-3.35%0.0068
(1871.875, 3819.857177734375]22.92%21.14%-1.78%0.0014
(3819.857177734375, 13787.5]23.40%26.11%2.71%0.0030
(13787.5, inf]5.31%9.55%4.24%0.0249
useratemore70_m6_num — PSI 0.0784
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 0.5]28.42%37.70%9.28%0.0262
(0.5, 1.5]11.65%15.06%3.41%0.0088
(1.5, 4.5]37.87%33.31%-4.55%0.0058
(4.5, 5.5]9.11%5.58%-3.53%0.0173
(5.5, inf]12.95%8.34%-4.61%0.0203
shd_pre_repay_amt_rate — PSI 0.0403
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 0.6238499879837036]33.20%41.50%8.30%0.0185
(0.6238499879837036, 0.7506499886512756]22.68%16.98%-5.70%0.0165
(0.7506499886512756, 0.8537499904632568]23.82%20.65%-3.17%0.0045
(0.8537499904632568, 0.9255499839782715]15.13%15.04%-0.09%0.0000
(0.9255499839782715, inf]5.17%5.83%0.66%0.0008
repay_amount_m6_all_shd_amt_pc — PSI 0.0409
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 93.98500061035156]34.83%41.73%6.90%0.0125
(93.98500061035156, 95.59499740600586]7.97%5.85%-2.12%0.0066
(95.59499740600586, 96.48500061035156]7.82%5.55%-2.27%0.0078
(96.48500061035156, 99.48500061035156]36.09%30.74%-5.35%0.0086
(99.48500061035156, inf]13.29%16.13%2.83%0.0055
cre_now_l30_used_amt_d_max — PSI 0.2379
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 0.6786499917507172]19.47%14.43%-5.03%0.0151
(0.6786499917507172, 0.7513499855995178]6.34%3.94%-2.41%0.0115
(0.7513499855995178, 0.7632499933242798]5.06%3.11%-1.95%0.0095
(0.7632499933242798, 0.9139499962329865]46.11%32.76%-13.35%0.0456
(0.9139499962329865, inf]23.02%45.76%22.74%0.1562
ps_int_amt_avg — PSI 0.1191
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 4.333299875259399]10.72%22.49%11.76%0.0871
(4.333299875259399, 13.080849647521973]32.11%24.75%-7.36%0.0192
(13.080849647521973, 15.02239990234375]5.37%4.48%-0.89%0.0016
(15.02239990234375, 38.24515151977539]32.32%27.00%-5.32%0.0096
(38.24515151977539, inf]19.48%21.30%1.81%0.0016
useratemore90maxcontimonsm12 — PSI 0.1030
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 0.5]24.43%37.30%12.87%0.0544
(0.5, 1.5]18.14%19.61%1.47%0.0012
(1.5, 4.5]38.75%31.06%-7.70%0.0170
(4.5, 7.5]12.22%7.37%-4.85%0.0246
(7.5, inf]6.46%4.67%-1.79%0.0058
repay_amount_all_shd_amt_pc — PSI 0.2346
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, -4999.0]5.67%20.00%14.33%0.1806
(-4999.0, 90.52499771118164]24.19%26.40%2.21%0.0019
(90.52499771118164, 97.21500015258789]42.27%30.51%-11.76%0.0384
(97.21500015258789, 98.58499908447266]18.34%13.67%-4.67%0.0137
(98.58499908447266, inf]9.52%9.42%-0.10%0.0000
min_apply_amt — PSI 0.0271
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 1135.0]8.90%12.49%3.59%0.0122
(1135.0, 2270.0]13.50%12.57%-0.93%0.0007
(2270.0, 4180.0]33.72%29.06%-4.66%0.0069
(4180.0, 7300.0]24.21%22.57%-1.65%0.0012
(7300.0, inf]19.66%23.31%3.65%0.0062
useratemore90_m3_num — PSI 0.0027
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 0.5]61.13%59.31%-1.82%0.0006
(0.5, 1.5]16.07%17.37%1.29%0.0010
(1.5, 2.5]10.45%11.47%1.01%0.0009
(2.5, inf]12.34%11.86%-0.49%0.0002
cre_now_l60_usedrate_m_max — PSI 0.2041
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 4.999999873689376e-05]13.38%30.20%16.82%0.1369
(4.999999873689376e-05, 0.039750000461936]6.91%7.79%0.88%0.0010
(0.039750000461936, 0.08044999837875366]8.09%8.63%0.54%0.0003
(0.08044999837875366, 0.5794999897480011]64.28%45.45%-18.84%0.0653
(0.5794999897480011, inf]7.33%7.94%0.60%0.0005
min_tnr — PSI 0.0040
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 1.5]7.06%5.87%-1.19%0.0022
(1.5, 5.0]53.21%52.11%-1.09%0.0002
(5.0, 8.5]7.91%8.84%0.93%0.0010
(8.5, inf]31.83%33.18%1.35%0.0006
loan_os_prcp_gt0_m12_num_pc — PSI 0.0370
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 57.73500061035156]8.67%14.16%5.50%0.0270
(57.73500061035156, 72.08000183105469]6.91%7.87%0.96%0.0012
(72.08000183105469, 76.01499938964844]5.15%5.32%0.17%0.0001
(76.01499938964844, 92.31500244140625]17.45%14.17%-3.28%0.0068
(92.31500244140625, inf]61.82%58.48%-3.34%0.0019
grade_limit_amount — PSI 0.2077
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 84500.0]5.72%1.94%-3.78%0.0409
(84500.0, 115500.0]24.73%12.51%-12.22%0.0832
(115500.0, 137500.0]5.34%3.27%-2.07%0.0101
(137500.0, 180500.0]16.10%14.00%-2.10%0.0029
(180500.0, inf]48.11%68.27%20.16%0.0705
end_loan_to_rep_loan_freq — PSI 0.0395
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 1003.5]75.31%76.96%1.66%0.0004
(1003.5, 1935.5]5.28%4.78%-0.50%0.0005
(1935.5, 2132.5]5.95%2.86%-3.09%0.0227
(2132.5, 2322.5]7.00%5.70%-1.30%0.0027
(2322.5, inf]6.46%9.70%3.24%0.0132
shd_pre_repay_cnt — PSI 0.0547
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 25.5]25.12%35.69%10.57%0.0371
(25.5, 91.5]30.28%24.81%-5.47%0.0109
(91.5, 237.5]19.79%16.93%-2.86%0.0045
(237.5, 362.5]6.67%6.19%-0.48%0.0004
(362.5, inf]18.14%16.38%-1.76%0.0018
last_clear_num_m1 — PSI 0.0046
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, 0.5]79.20%81.89%2.69%0.0009
(0.5, 1.5]12.91%11.22%-1.69%0.0024
(1.5, inf]7.89%6.88%-1.00%0.0014
setl_rem_prcp_paym_last_month_lo — PSI 0.1082
分栏开发基准比例PSI时点样本 比例变更贡献
(-inf, -4999.0]7.62%18.39%10.77%0.0949
(-4999.0, inf]92.38%81.61%-10.77%0.0133

注:PSI ≤ 0.1 视为稳定,0.1~0.25 需关注,> 0.25 不稳定。

12 单变量分析

注:统计所有入模变量的单变量分析(开发样本口径)。预留验证/总体/OOT 口径的分组明细待扩展。

f_ps_od_days_max
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 0.5]2042253.63%18144227811.15%7.96
2(0.5, 1.5]844175.80%80024395.20%18.23
3(1.5, 2.5]342084.78%32751454.24%22.59
4(2.5, 4.5]359094.20%34531373.82%25.20
5(4.5, inf]2207100.00%20561516.84%13.62
合计380803493031508.27%11.09
2026-11-02T16:28:03.241041 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(11.15% → 11.15% → 6.84%),如需严格单调可考虑调整切点或合并分箱。IV=0.1937,该变量区分度尚可。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 4→5 箱有局部反弹:Bad Rate 由 3.82% 回升至 6.84%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 0.5])为坏率最高区间:坏率高达 11.15%,可作为拒绝或重点风控策略的阈值参考

suc_apprv_num_m1
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 0.5]3016979.23%2728728829.55%9.47
2(0.5, 1.5]478791.80%45911964.09%23.42
3(1.5, inf]3124100.00%3052722.30%42.39
合计380803493031508.27%11.09
2026-11-02T16:28:03.375590 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调下降(9.55% → 2.30%),WOE 单调上升(-0.1580 → 1.3409),符合风控分箱预期。IV=0.1594,该变量区分度尚可。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 2 箱 WOE<0(风险高于平均)

第 1 箱((-inf, 0.5])为坏率最高区间:坏率高达 9.55%,可作为拒绝或重点风控策略的阈值参考

apprv_amt_day
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 11.0]3034479.68%2745228929.53%9.49
2(11.0, 1000.5]250686.27%2421853.39%28.48
3(1000.5, 5995.0]330494.94%3220842.54%38.33
4(5995.0, inf]1926100.00%1837894.62%20.64
合计380803493031508.27%11.09
2026-11-02T16:28:03.507745 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(9.53% → 9.53% → 4.62%),如需严格单调可考虑调整切点或合并分箱。IV=0.1569,该变量区分度尚可。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 3→4 箱有局部反弹:Bad Rate 由 2.54% 回升至 4.62%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 11.0])为坏率最高区间:坏率高达 9.53%,可作为拒绝或重点风控策略的阈值参考

t_ps_od_days_max
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 0.5]2063354.18%18517211610.26%8.75
2(0.5, 2.5]1109783.32%105245735.16%18.37
3(2.5, 3.5]204888.70%1977713.47%27.85
4(3.5, 5.5]219794.47%20751225.55%17.01
5(5.5, inf]2105100.00%183726812.73%6.85
合计380803493031508.27%11.09
2026-11-02T16:28:03.639030 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(10.26% → 12.73% → 12.73%),如需严格单调可考虑调整切点或合并分箱。IV=0.1496,该变量区分度尚可。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 1→2 箱有局部反弹:Bad Rate 由 10.26% 回落至 5.16%,如需严格单调可考虑合并这两箱

第 5 箱((5.5, inf])为坏率最高区间:坏率高达 12.73%,可作为拒绝或重点风控策略的阈值参考

cre_now_l60_used_amt_m_min
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 1.7799999713897705]2820174.09%2545827439.73%9.28
2(1.7799999713897705, 869.5950012207031]247680.59%23681084.36%21.93
3(869.5950012207031, 1999.7100219726562]206086.00%2009512.48%39.39
4(1999.7100219726562, 6462.93505859375]312594.21%30091163.71%25.94
5(6462.93505859375, inf]2203100.00%20711325.99%15.69
合计380653491531508.28%11.08
2026-11-02T16:28:03.779692 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(9.73% → 9.73% → 5.99%),如需严格单调可考虑调整切点或合并分箱。IV=0.1485,该变量区分度尚可。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 3→4 箱有局部反弹:Bad Rate 由 2.48% 回升至 3.71%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 1.7799999713897705])为坏率最高区间:坏率高达 9.73%,可作为拒绝或重点风控策略的阈值参考

cre_l60_used_limit_rate_min
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 0.019700000062584877]444111.67%43061353.04%31.90
2(0.019700000062584877, 0.09174999967217445]281319.06%26561575.58%16.92
3(0.09174999967217445, 0.2787500023841858]732338.30%68075167.05%13.19
4(0.2787500023841858, 0.9161500036716461]1889987.94%1715717429.22%9.85
5(0.9161500036716461, inf]4589100.00%398960013.07%6.65
合计380653491531508.28%11.08
2026-11-02T16:28:03.990764 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调上升(3.04% → 13.07%),WOE 单调下降(1.0570 → -0.5111),符合风控分箱预期。IV=0.1478,该变量区分度尚可。

WOE 在第 3、4 箱之间穿越 0 线:前 3 箱 WOE>0(客户质量优于整体),后 2 箱 WOE<0(风险高于平均)

第 5 箱((0.9161500036716461, inf])为坏率最高区间:坏率高达 13.07%,可作为拒绝或重点风控策略的阈值参考

s_ps_od_days_max
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 0.5]2061054.12%18430218010.58%8.45
2(0.5, 1.5]795175.00%75284235.32%17.80
3(1.5, 2.5]346584.10%33031624.68%20.39
4(2.5, 4.5]349493.28%33261684.81%19.80
5(4.5, inf]2560100.00%23432178.48%10.80
合计380803493031508.27%11.09
2026-11-02T16:28:04.143276 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(10.58% → 10.58% → 8.48%),如需严格单调可考虑调整切点或合并分箱。IV=0.1336,该变量区分度尚可。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 3→4 箱有局部反弹:Bad Rate 由 4.68% 回升至 4.81%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 0.5])为坏率最高区间:坏率高达 10.58%,可作为拒绝或重点风控策略的阈值参考

cre_l90_used_limit_amt_min
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 21.640000343322754]561814.76%54082103.74%25.75
2(21.640000343322754, 748.7099914550781]807635.98%75265506.81%13.68
3(748.7099914550781, 8516.22998046875]2017288.97%1835418189.01%10.10
4(8516.22998046875, 14597.384765625]226794.92%199027712.22%7.18
5(14597.384765625, inf]1932100.00%163729515.27%5.55
合计380653491531508.28%11.08
2026-11-02T16:28:04.294518 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调上升(3.74% → 15.27%),WOE 单调下降(0.8430 → -0.6919),符合风控分箱预期。IV=0.1336,该变量区分度尚可。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 5 箱((14597.384765625, inf])为坏率最高区间:坏率高达 15.27%,可作为拒绝或重点风控策略的阈值参考

loan_os_prcp_credit_amount_m3
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, -4999.0]36779.66%329937810.28%8.73
2(-4999.0, 35.125]831531.51%79333824.59%20.77
3(35.125, 100.47500228881836]1911681.73%1761015067.88%11.69
4(100.47500228881836, 124.34499740600586]271788.87%241230511.23%7.91
5(124.34499740600586, inf]4235100.00%365757813.65%6.33
合计380603491131498.27%11.09
2026-11-02T16:28:04.434036 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(10.28% → 13.65% → 13.65%),如需严格单调可考虑调整切点或合并分箱。IV=0.1276,该变量区分度尚可。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 1→2 箱有局部反弹:Bad Rate 由 10.28% 回落至 4.59%,如需严格单调可考虑合并这两箱

第 5 箱((124.34499740600586, inf])为坏率最高区间:坏率高达 13.65%,可作为拒绝或重点风控策略的阈值参考

apprv_num_m1
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 0.5]2815273.93%2542527279.69%9.32
2(0.5, 1.5]535087.98%51072434.54%21.02
3(1.5, 2.5]198993.20%1914753.77%25.52
4(2.5, inf]2589100.00%24841054.06%23.66
合计380803493031508.27%11.09
2026-11-02T16:28:04.565352 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(9.69% → 9.69% → 4.06%),如需严格单调可考虑调整切点或合并分箱。IV=0.1225,该变量区分度尚可。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 3→4 箱有局部反弹:Bad Rate 由 3.77% 回升至 4.06%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 0.5])为坏率最高区间:坏率高达 9.69%,可作为拒绝或重点风控策略的阈值参考

cre_now_l60_limit_left_d_max
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, -0.013299999758601189]22535.94%199925411.27%7.87
2(-0.013299999758601189, 0.3223000019788742]393516.33%37202155.46%17.30
3(0.3223000019788742, 0.6971500217914581]239722.65%22881094.55%20.99
4(0.6971500217914581, 0.9999499917030334]289030.27%2809812.80%34.68
5(0.9999499917030334, inf]26429100.00%2396024699.34%9.70
合计379043477631288.25%11.12
2026-11-02T16:28:04.696257 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(11.27% → 11.27% → 9.34%),如需严格单调可考虑调整切点或合并分箱。IV=0.1208,该变量区分度尚可。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 4→5 箱有局部反弹:Bad Rate 由 2.80% 回升至 9.34%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, -0.013299999758601189])为坏率最高区间:坏率高达 11.27%,可作为拒绝或重点风控策略的阈值参考

available_limit_amt
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 0.03499999921768904]1238532.52%10918146711.85%7.44
2(0.03499999921768904, 468.864990234375]452344.40%42472766.10%15.39
3(468.864990234375, 4307.300048828125]1199875.91%111488507.08%13.12
4(4307.300048828125, 27316.265625]725094.95%67884626.37%14.69
5(27316.265625, inf]1924100.00%1829954.94%19.25
合计380803493031508.27%11.09
2026-11-02T16:28:04.837480 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(11.85% → 11.85% → 4.94%),如需严格单调可考虑调整切点或合并分箱。IV=0.1061,该变量区分度尚可。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 2→3 箱有局部反弹:Bad Rate 由 6.10% 回升至 7.08%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 0.03499999921768904])为坏率最高区间:坏率高达 11.85%,可作为拒绝或重点风控策略的阈值参考

loan_os_prcp_credit_amount_m6
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 4.129999876022339]19175.04%17931246.47%14.46
2(4.129999876022339, 33.795000076293945]621021.35%59412694.33%22.09
3(33.795000076293945, 48.94499969482422]443733.01%41412966.67%13.99
4(48.94499969482422, 107.59499740600586]1873582.24%1707416618.87%10.28
5(107.59499740600586, inf]6761100.00%596279911.82%7.46
合计380603491131498.27%11.09
2026-11-02T16:28:04.980737 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(6.47% → 11.82% → 11.82%),如需严格单调可考虑调整切点或合并分箱。IV=0.1031,该变量区分度尚可。

WOE 在第 3、4 箱之间穿越 0 线:前 3 箱 WOE>0(客户质量优于整体),后 2 箱 WOE<0(风险高于平均)

第 1→2 箱有局部反弹:Bad Rate 由 6.47% 回落至 4.33%,如需严格单调可考虑合并这两箱

第 5 箱((107.59499740600586, inf])为坏率最高区间:坏率高达 11.82%,可作为拒绝或重点风控策略的阈值参考

cre_now_l30_used_amt_d_min
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 1.0000500082969666]3121186.91%2827329389.41%9.62
2(1.0000500082969666, 1.382200002670288]193392.29%1860733.78%25.48
3(1.382200002670288, inf]2768100.00%2687812.93%33.17
合计359123282030928.61%10.61
2026-11-02T16:28:05.117308 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调下降(9.41% → 2.93%),WOE 单调上升(-0.0980 → 1.1395),符合风控分箱预期。IV=0.1011,该变量区分度尚可。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 2 箱 WOE<0(风险高于平均)

第 1 箱((-inf, 1.0000500082969666])为坏率最高区间:坏率高达 9.41%,可作为拒绝或重点风控策略的阈值参考

loan_os_prcp_credit_amount_m12
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 16.18000030517578]31648.31%30581063.35%28.85
2(16.18000030517578, 44.21000099182129]603324.16%56713626.00%15.67
3(44.21000099182129, 77.875]1445362.14%1322312308.51%10.75
4(77.875, 95.76499938964844]663579.57%60086279.45%9.58
5(95.76499938964844, inf]7775100.00%695182410.60%8.44
合计380603491131498.27%11.09
2026-11-02T16:28:05.253773 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调上升(3.35% → 10.60%),WOE 单调下降(0.9564 → -0.2732),符合风控分箱预期。IV=0.0894,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 5 箱((95.76499938964844, inf])为坏率最高区间:坏率高达 10.60%,可作为拒绝或重点风控策略的阈值参考

cre_now_l60_used_amt_d_min
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 1.000950038433075]2794381.74%2520527389.80%9.21
2(1.000950038433075, 1.4549500346183777]235688.63%22351215.14%18.47
3(1.4549500346183777, 2.305649995803833]186094.07%1789713.82%25.20
4(2.305649995803833, inf]2026100.00%19261004.94%19.26
合计341853115530308.86%10.28
2026-11-02T16:28:05.392269 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(9.80% → 9.80% → 4.94%),如需严格单调可考虑调整切点或合并分箱。IV=0.0776,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 3→4 箱有局部反弹:Bad Rate 由 3.82% 回升至 4.94%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 1.000950038433075])为坏率最高区间:坏率高达 9.80%,可作为拒绝或重点风控策略的阈值参考

loan_os_prcp_gt0_m3_num_pc
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 68.16500091552734]33548.81%32521023.04%31.88
2(68.16500091552734, inf]34706100.00%3165930478.78%10.39
合计380603491131498.27%11.09
2026-11-02T16:28:05.587968 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调上升(3.04% → 8.78%),WOE 单调下降(1.0563 → -0.0649),符合风控分箱预期。IV=0.0681,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 1 箱 WOE<0(风险高于平均)

第 2 箱((68.16500091552734, inf])为坏率最高区间:坏率高达 8.78%,可作为拒绝或重点风控策略的阈值参考

cre_now_l90_limit_left_d_min
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, -0.6798500120639801]27007.62%25101907.04%13.21
2(-0.6798500120639801, 0.1424499973654747]288615.75%26242629.08%10.02
3(0.1424499973654747, 0.9937500059604645]410327.33%356354013.16%6.60
4(0.9937500059604645, 1.1872499585151672]791249.64%74344786.04%15.55
5(1.1872499585151672, inf]17855100.00%1634715088.45%10.84
合计354563247829788.40%10.91
2026-11-02T16:28:05.714432 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(7.04% → 13.16% → 8.45%),如需严格单调可考虑调整切点或合并分箱。IV=0.0635,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 3→4 箱有局部反弹:Bad Rate 由 13.16% 回落至 6.04%,如需严格单调可考虑合并这两箱

第 3 箱((0.1424499973654747, 0.9937500059604645])为坏率最高区间:坏率高达 13.16%,可作为拒绝或重点风控策略的阈值参考

useratemore50maxcontimonsm12
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 0.5]407610.71%38891874.59%20.80
2(0.5, 1.5]200315.97%18701336.64%14.06
3(1.5, 2.5]228621.98%21651215.29%17.89
4(2.5, 4.5]589337.46%54014928.35%10.98
5(4.5, inf]23802100.00%2158622169.31%9.74
合计380603491131498.27%11.09
2026-11-02T16:28:05.853409 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(4.59% → 9.31% → 9.31%),如需严格单调可考虑调整切点或合并分箱。IV=0.0578,该变量区分度偏弱但可用。

WOE 在第 3、4 箱之间穿越 0 线:前 3 箱 WOE>0(客户质量优于整体),后 2 箱 WOE<0(风险高于平均)

第 2→3 箱有局部反弹:Bad Rate 由 6.64% 回落至 5.29%,如需严格单调可考虑合并这两箱

第 5 箱((4.5, inf])为坏率最高区间:坏率高达 9.31%,可作为拒绝或重点风控策略的阈值参考

cre_now_p1m_used_rate_m
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, -0.13974999636411667]513713.61%48073306.42%14.57
2(-0.13974999636411667, -4.999999873689376e-05]215119.31%2063884.09%23.44
3(-4.999999873689376e-05, 0.0002500000045984052]190324.36%165724612.93%6.74
4(0.0002500000045984052, 0.04154999926686287]429435.74%40022926.80%13.71
5(0.04154999926686287, inf]24249100.00%2207621738.96%10.16
合计377343460531298.29%11.06
2026-11-02T16:28:05.987982 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(6.42% → 12.93% → 8.96%),如需严格单调可考虑调整切点或合并分箱。IV=0.0577,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 1→2 箱有局部反弹:Bad Rate 由 6.42% 回落至 4.09%,如需严格单调可考虑合并这两箱

第 3 箱((-4.999999873689376e-05, 0.0002500000045984052])为坏率最高区间:坏率高达 12.93%,可作为拒绝或重点风控策略的阈值参考

ps_int_amt_sum
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 209.51499938964844]31738.33%30021715.39%17.56
2(209.51499938964844, 815.2900085449219]1035635.53%96177397.14%13.01
3(815.2900085449219, 1891.7449951171875]1042962.91%96138167.82%11.78
4(1891.7449951171875, 10533.93994140625]1131692.63%1027610409.19%9.88
5(10533.93994140625, inf]2806100.00%242238413.69%6.31
合计380803493031508.27%11.09
2026-11-02T16:28:06.124967 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调上升(5.39% → 13.69%),WOE 单调下降(0.4594 → -0.5642),符合风控分箱预期。IV=0.0558,该变量区分度偏弱但可用。

WOE 在第 3、4 箱之间穿越 0 线:前 3 箱 WOE>0(客户质量优于整体),后 2 箱 WOE<0(风险高于平均)

第 5 箱((10533.93994140625, inf])为坏率最高区间:坏率高达 13.69%,可作为拒绝或重点风控策略的阈值参考

cre_now_l60_used_amt_m_max
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 83.30500030517578]514813.52%49042444.74%20.10
2(83.30500030517578, 1001.9599914550781]1693758.02%1556813698.08%11.37
3(1001.9599914550781, 2003.010009765625]715176.81%65156368.89%10.24
4(2003.010009765625, 2696.469970703125]231882.90%205426411.39%7.78
5(2696.469970703125, inf]6511100.00%58746379.78%9.22
合计380653491531508.28%11.08
2026-11-02T16:28:06.263345 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(4.74% → 11.39% → 9.78%),如需严格单调可考虑调整切点或合并分箱。IV=0.0541,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 4→5 箱有局部反弹:Bad Rate 由 11.39% 回落至 9.78%,如需严格单调可考虑合并这两箱

第 4 箱((2003.010009765625, 2696.469970703125])为坏率最高区间:坏率高达 11.39%,可作为拒绝或重点风控策略的阈值参考

cre_l30_l90_used_amt_m_min
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 0.6299999952316284]2985678.43%2718526718.95%10.18
2(0.6299999952316284, 1107.5999755859375]285885.94%27421164.06%23.64
3(1107.5999755859375, 5082.195068359375]325894.50%30821765.40%17.51
4(5082.195068359375, inf]2093100.00%19061878.93%10.19
合计380653491531508.28%11.08
2026-11-02T16:28:06.394717 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(8.95% → 8.95% → 8.93%),如需严格单调可考虑调整切点或合并分箱。IV=0.0527,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 2→3 箱有局部反弹:Bad Rate 由 4.06% 回升至 5.40%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 0.6299999952316284])为坏率最高区间:坏率高达 8.95%,可作为拒绝或重点风控策略的阈值参考

avg_tnr 平均履约月数
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 3.5]24366.40%2340963.94%24.38
2(3.5, 5.5]367216.04%34422306.26%14.97
3(5.5, 6.5]266023.03%24721887.07%13.15
4(6.5, 10.5]1267156.30%1164510268.10%11.35
5(10.5, inf]16641100.00%1503116109.67%9.34
合计380803493031508.27%11.09
2026-11-02T16:28:06.524770 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调上升(3.94% → 9.67%),WOE 单调下降(0.7876 → -0.1721),符合风控分箱预期。IV=0.0524,该变量区分度偏弱但可用。

WOE 在第 4、5 箱之间穿越 0 线:前 4 箱 WOE>0(客户质量优于整体),后 1 箱 WOE<0(风险高于平均)

第 5 箱((10.5, inf])为坏率最高区间:坏率高达 9.67%,可作为拒绝或重点风控策略的阈值参考

cre_now_l30_limit_left_d_min
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 0.26759999990463257]26137.16%23772369.03%10.07
2(0.26759999990463257, 0.997050017118454]352316.82%306645712.97%6.71
3(0.997050017118454, 1.1242499947547913]1377354.58%1277010037.28%12.73
4(1.1242499947547913, 1.7257500290870667]1162786.46%1058510428.96%10.16
5(1.7257500290870667, inf]4939100.00%46203196.46%14.48
合计364753341830578.38%10.93
2026-11-02T16:28:06.663832 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(9.03% → 12.97% → 6.46%),如需严格单调可考虑调整切点或合并分箱。IV=0.0482,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 1→2 箱有局部反弹:Bad Rate 由 9.03% 回升至 12.97%,如需严格单调可考虑合并这两箱

第 2 箱((0.26759999990463257, 0.997050017118454])为坏率最高区间:坏率高达 12.97%,可作为拒绝或重点风控策略的阈值参考

max_loan_os_prcp_m12
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 1747.2849731445312]410110.78%38602415.88%16.02
2(1747.2849731445312, 2010.25]280618.15%25622448.70%10.50
3(2010.25, 13572.5849609375]2248077.21%2076617147.62%12.12
4(13572.5849609375, 20848.58984375]361986.72%32783419.42%9.61
5(20848.58984375, inf]5054100.00%444560912.05%7.30
合计380603491131498.27%11.09
2026-11-02T16:28:06.814008 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(5.88% → 12.05% → 12.05%),如需严格单调可考虑调整切点或合并分箱。IV=0.0469,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 2→3 箱有局部反弹:Bad Rate 由 8.70% 回落至 7.62%,如需严格单调可考虑合并这两箱

第 5 箱((20848.58984375, inf])为坏率最高区间:坏率高达 12.05%,可作为拒绝或重点风控策略的阈值参考

suc_apprv_num_m6
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 0.5]1327234.85%1196013129.89%9.12
2(0.5, 1.5]873957.80%79657748.86%10.29
3(1.5, 2.5]477070.33%44383326.96%13.37
4(2.5, 3.5]304278.32%28022407.89%11.68
5(3.5, inf]8257100.00%77654925.96%15.78
合计380803493031508.27%11.09
2026-11-02T16:28:06.959128 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(9.89% → 9.89% → 5.96%),如需严格单调可考虑调整切点或合并分箱。IV=0.0434,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 3→4 箱有局部反弹:Bad Rate 由 6.96% 回升至 7.89%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 0.5])为坏率最高区间:坏率高达 9.89%,可作为拒绝或重点风控策略的阈值参考

useratemore70monthsubcurmonth
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 0.5]1364535.85%1244412018.80%10.36
2(0.5, 2.5]658653.16%59736139.31%9.74
3(2.5, 4.5]519866.81%46855139.87%9.13
4(4.5, 9.5]603782.67%55644737.83%11.76
5(9.5, inf]6594100.00%62453495.29%17.89
合计380603491131498.27%11.09
2026-11-02T16:28:07.174281 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(8.80% → 9.87% → 5.29%),如需严格单调可考虑调整切点或合并分箱。IV=0.0434,该变量区分度偏弱但可用。

WOE 在第 3、4 箱之间穿越 0 线:前 3 箱 WOE>0(客户质量优于整体),后 2 箱 WOE<0(风险高于平均)

第 1→2 箱有局部反弹:Bad Rate 由 8.80% 回升至 9.31%,如需严格单调可考虑合并这两箱

第 3 箱((2.5, 4.5])为坏率最高区间:坏率高达 9.87%,可作为拒绝或重点风控策略的阈值参考

setl_rem_prcp_paym_m3_num
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 0.5]3209484.32%2924028548.89%10.25
2(0.5, inf]5966100.00%56712954.94%19.22
合计380603491131498.27%11.09
2026-11-02T16:28:07.303797 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调下降(8.89% → 4.94%),WOE 单调上升(-0.0789 → 0.5504),符合风控分箱预期。IV=0.0433,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 1 箱 WOE<0(风险高于平均)

第 1 箱((-inf, 0.5])为坏率最高区间:坏率高达 8.89%,可作为拒绝或重点风控策略的阈值参考

orig_prcp_avg_m12
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 1031.9722290039062]1149830.19%107357636.64%14.07
2(1031.9722290039062, 1871.875]692548.38%64055207.51%12.32
3(1871.875, 3819.857177734375]872771.30%79937348.41%10.89
4(3819.857177734375, 13787.5]890994.69%80418689.74%9.26
5(13787.5, inf]2021100.00%175626513.11%6.63
合计380803493031508.27%11.09
2026-11-02T16:28:07.438682 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调上升(6.64% → 13.11%),WOE 单调下降(0.2381 → -0.5149),符合风控分箱预期。IV=0.0431,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 5 箱((13787.5, inf])为坏率最高区间:坏率高达 13.11%,可作为拒绝或重点风控策略的阈值参考

useratemore70_m6_num
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 0.5]1081728.42%101466716.20%15.12
2(0.5, 1.5]443540.07%41153207.22%12.86
3(1.5, 4.5]1441277.94%1310413089.08%10.02
4(4.5, 5.5]346887.05%31323369.69%9.32
5(5.5, inf]4928100.00%441451410.43%8.59
合计380603491131498.27%11.09
2026-11-02T16:28:07.575185 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调上升(6.20% → 10.43%),WOE 单调下降(0.3103 → -0.2554),符合风控分箱预期。IV=0.0429,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 5 箱((5.5, inf])为坏率最高区间:坏率高达 10.43%,可作为拒绝或重点风控策略的阈值参考

shd_pre_repay_amt_rate
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 0.6238499879837036]1263733.20%118148236.51%14.35
2(0.6238499879837036, 0.7506499886512756]863155.88%79446877.96%11.56
3(0.7506499886512756, 0.8537499904632568]906679.70%82188489.35%9.69
4(0.8537499904632568, 0.9255499839782715]575994.83%513162810.90%8.17
5(0.9255499839782715, inf]1966100.00%18031638.29%11.06
合计380593491031498.27%11.09
2026-11-02T16:28:07.705081 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(6.51% → 10.90% → 8.29%),如需严格单调可考虑调整切点或合并分箱。IV=0.0409,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 4→5 箱有局部反弹:Bad Rate 由 10.90% 回落至 8.29%,如需严格单调可考虑合并这两箱

第 4 箱((0.8537499904632568, 0.9255499839782715])为坏率最高区间:坏率高达 10.90%,可作为拒绝或重点风控策略的阈值参考

repay_amount_m6_all_shd_amt_pc
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 93.98500061035156]1325734.83%1194613119.89%9.11
2(93.98500061035156, 95.59499740600586]303342.80%28122217.29%12.72
3(95.59499740600586, 96.48500061035156]297550.62%26942819.45%9.59
4(96.48500061035156, 99.48500061035156]1373586.71%1268710487.63%12.11
5(99.48500061035156, inf]5060100.00%47722885.69%16.57
合计380603491131498.27%11.09
2026-11-02T16:28:07.842387 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(9.89% → 9.89% → 5.69%),如需严格单调可考虑调整切点或合并分箱。IV=0.0386,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 2→3 箱有局部反弹:Bad Rate 由 7.29% 回升至 9.45%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 93.98500061035156])为坏率最高区间:坏率高达 9.89%,可作为拒绝或重点风控策略的阈值参考

cre_now_l30_used_amt_d_max
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 0.6786499917507172]737619.47%68984786.48%14.43
2(0.6786499917507172, 0.7513499855995178]240425.81%21862189.07%10.03
3(0.7513499855995178, 0.7632499933242798]191630.87%17791377.15%12.99
4(0.7632499933242798, 0.9139499962329865]1747176.98%1577117009.73%9.28
5(0.9139499962329865, inf]8724100.00%81106147.04%13.21
合计378913474431478.31%11.04
2026-11-02T16:28:07.978352 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(6.48% → 9.73% → 7.04%),如需严格单调可考虑调整切点或合并分箱。IV=0.0362,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 2→3 箱有局部反弹:Bad Rate 由 9.07% 回落至 7.15%,如需严格单调可考虑合并这两箱

第 4 箱((0.7632499933242798, 0.9139499962329865])为坏率最高区间:坏率高达 9.73%,可作为拒绝或重点风控策略的阈值参考

ps_int_amt_avg
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 4.333299875259399]408310.72%38572265.54%17.07
2(4.333299875259399, 13.080849647521973]1222742.83%113149137.47%12.39
3(13.080849647521973, 15.02239990234375]204548.20%18412049.98%9.02
4(15.02239990234375, 38.24515151977539]1230680.52%1126510418.46%10.82
5(38.24515151977539, inf]7419100.00%665376610.32%8.69
合计380803493031508.27%11.09
2026-11-02T16:28:08.119827 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(5.54% → 10.32% → 10.32%),如需严格单调可考虑调整切点或合并分箱。IV=0.0360,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 3→4 箱有局部反弹:Bad Rate 由 9.98% 回落至 8.46%,如需严格单调可考虑合并这两箱

第 5 箱((38.24515151977539, inf])为坏率最高区间:坏率高达 10.32%,可作为拒绝或重点风控策略的阈值参考

useratemore90maxcontimonsm12
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 0.5]929824.43%87235756.18%15.17
2(0.5, 1.5]690342.57%63495548.03%11.46
3(1.5, 4.5]1475081.32%1346112898.74%10.44
4(4.5, 7.5]465293.54%41884649.97%9.03
5(7.5, inf]2457100.00%219026710.87%8.20
合计380603491131498.27%11.09
2026-11-02T16:28:08.257068 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调上升(6.18% → 10.87%),WOE 单调下降(0.3136 → -0.3013),符合风控分箱预期。IV=0.0350,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 5 箱((7.5, inf])为坏率最高区间:坏率高达 10.87%,可作为拒绝或重点风控策略的阈值参考

repay_amount_all_shd_amt_pc
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, -4999.0]21595.67%20171426.58%14.20
2(-4999.0, 90.52499771118164]920729.86%83069019.79%9.22
3(90.52499771118164, 97.21500015258789]1608972.14%1468914008.70%10.49
4(97.21500015258789, 98.58499908447266]698290.48%64825007.16%12.96
5(98.58499908447266, inf]3623100.00%34172065.69%16.59
合计380603491131498.27%11.09
2026-11-02T16:28:08.388945 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(6.58% → 9.79% → 5.69%),如需严格单调可考虑调整切点或合并分箱。IV=0.0306,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 1→2 箱有局部反弹:Bad Rate 由 6.58% 回升至 9.79%,如需严格单调可考虑合并这两箱

第 2 箱((-4999.0, 90.52499771118164])为坏率最高区间:坏率高达 9.79%,可作为拒绝或重点风控策略的阈值参考

min_apply_amt
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 1135.0]33888.90%301037811.16%7.96
2(1135.0, 2270.0]514222.40%47344087.93%11.60
3(2270.0, 4180.0]1284156.12%1167911629.05%10.05
4(4180.0, 7300.0]922180.34%84907317.93%11.61
5(7300.0, inf]7488100.00%70174716.29%14.90
合计380803493031508.27%11.09
2026-11-02T16:28:08.523287 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(11.16% → 11.16% → 6.29%),如需严格单调可考虑调整切点或合并分箱。IV=0.0305,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 2→3 箱有局部反弹:Bad Rate 由 7.93% 回升至 9.05%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 1135.0])为坏率最高区间:坏率高达 11.16%,可作为拒绝或重点风控策略的阈值参考

useratemore90_m3_num
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 0.5]2326661.13%2150117657.59%12.18
2(0.5, 1.5]611777.20%56464717.70%11.99
3(1.5, 2.5]397987.66%36233568.95%10.18
4(2.5, inf]4698100.00%414155711.86%7.43
合计380603491131498.27%11.09
2026-11-02T16:28:08.649558 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调上升(7.59% → 11.86%),WOE 单调下降(0.0942 → -0.3996),符合风控分箱预期。IV=0.0302,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 2 箱 WOE<0(风险高于平均)

第 4 箱((2.5, inf])为坏率最高区间:坏率高达 11.86%,可作为拒绝或重点风控策略的阈值参考

cre_now_l60_usedrate_m_max
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 4.999999873689376e-05]509413.38%47493456.77%13.77
2(4.999999873689376e-05, 0.039750000461936]263220.30%24991335.05%18.79
3(0.039750000461936, 0.08044999837875366]307928.39%28462337.57%12.21
4(0.08044999837875366, 0.5794999897480011]2446992.67%2224522249.09%10.00
5(0.5794999897480011, inf]2791100.00%25762157.70%11.98
合计380653491531508.28%11.08
2026-11-02T16:28:08.860245 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(6.77% → 9.09% → 7.70%),如需严格单调可考虑调整切点或合并分箱。IV=0.0295,该变量区分度偏弱但可用。

WOE 在第 3、4 箱之间穿越 0 线:前 3 箱 WOE>0(客户质量优于整体),后 2 箱 WOE<0(风险高于平均)

第 1→2 箱有局部反弹:Bad Rate 由 6.77% 回落至 5.05%,如需严格单调可考虑合并这两箱

第 4 箱((0.08044999837875366, 0.5794999897480011])为坏率最高区间:坏率高达 9.09%,可作为拒绝或重点风控策略的阈值参考

min_tnr
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 1.5]26887.06%24342549.45%9.58
2(1.5, 5.0]2026160.27%1882614357.08%13.12
3(5.0, 8.5]301268.17%27372759.13%9.95
4(8.5, inf]12119100.00%1093311869.79%9.22
合计380803493031508.27%11.09
2026-11-02T16:28:09.001558 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(9.45% → 9.79% → 9.79%),如需严格单调可考虑调整切点或合并分箱。IV=0.0283,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 1→2 箱有局部反弹:Bad Rate 由 9.45% 回落至 7.08%,如需严格单调可考虑合并这两箱

第 4 箱((8.5, inf])为坏率最高区间:坏率高达 9.79%,可作为拒绝或重点风控策略的阈值参考

loan_os_prcp_gt0_m12_num_pc
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 57.73500061035156]32988.67%31381604.85%19.61
2(57.73500061035156, 72.08000183105469]263115.58%24172148.13%11.29
3(72.08000183105469, 76.01499938964844]196120.73%17841779.03%10.08
4(76.01499938964844, 92.31500244140625]664038.18%60945468.22%11.16
5(92.31500244140625, inf]23530100.00%2147820528.72%10.47
合计380603491131498.27%11.09
2026-11-02T16:28:09.135908 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(4.85% → 9.03% → 8.72%),如需严格单调可考虑调整切点或合并分箱。IV=0.0249,该变量区分度偏弱但可用。

WOE 在第 2、3 箱之间穿越 0 线:前 2 箱 WOE>0(客户质量优于整体),后 3 箱 WOE<0(风险高于平均)

第 3→4 箱有局部反弹:Bad Rate 由 9.03% 回落至 8.22%,如需严格单调可考虑合并这两箱

第 3 箱((72.08000183105469, 76.01499938964844])为坏率最高区间:坏率高达 9.03%,可作为拒绝或重点风控策略的阈值参考

grade_limit_amount
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 84500.0]21785.72%194523310.70%8.35
2(84500.0, 115500.0]941630.45%86567608.07%11.39
3(115500.0, 137500.0]203335.79%19001336.54%14.29
4(137500.0, 180500.0]613251.89%550163110.29%8.72
5(180500.0, inf]18321100.00%1692813937.60%12.15
合计380803493031508.27%11.09
2026-11-02T16:28:09.270778 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(10.70% → 10.70% → 7.60%),如需严格单调可考虑调整切点或合并分箱。IV=0.0226,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 3→4 箱有局部反弹:Bad Rate 由 6.54% 回升至 10.29%,如需严格单调可考虑合并这两箱

第 1 箱((-inf, 84500.0])为坏率最高区间:坏率高达 10.70%,可作为拒绝或重点风控策略的阈值参考

end_loan_to_rep_loan_freq
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 1003.5]2867775.31%2641922587.87%11.70
2(1003.5, 1935.5]201280.59%18191939.59%9.42
3(1935.5, 2132.5]226686.54%198428212.44%7.04
4(2132.5, 2322.5]266693.54%24172499.34%9.71
5(2322.5, inf]2459100.00%22911686.83%13.64
合计380803493031508.27%11.09
2026-11-02T16:28:09.403132 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(7.87% → 12.44% → 6.83%),如需严格单调可考虑调整切点或合并分箱。IV=0.0224,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 1→2 箱有局部反弹:Bad Rate 由 7.87% 回升至 9.59%,如需严格单调可考虑合并这两箱

第 3 箱((1935.5, 2132.5])为坏率最高区间:坏率高达 12.44%,可作为拒绝或重点风控策略的阈值参考

shd_pre_repay_cnt
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 25.5]956025.12%88766847.15%12.98
2(25.5, 91.5]1152655.40%105489788.49%10.79
3(91.5, 237.5]753175.19%69775547.36%12.59
4(237.5, 362.5]253781.86%23142238.79%10.38
5(362.5, inf]6905100.00%619571010.28%8.73
合计380593491031498.27%11.09
2026-11-02T16:28:09.557508 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性欠佳:Bad Rate 随分栏存在波动(7.15% → 10.28% → 10.28%),如需严格单调可考虑调整切点或合并分箱。IV=0.0209,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 4 箱 WOE<0(风险高于平均)

第 2→3 箱有局部反弹:Bad Rate 由 8.49% 回落至 7.36%,如需严格单调可考虑合并这两箱

第 5 箱((362.5, inf])为坏率最高区间:坏率高达 10.28%,可作为拒绝或重点风控策略的阈值参考

last_clear_num_m1
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, 0.5]3016179.20%2750926528.79%10.37
2(0.5, 1.5]491692.11%45943226.55%14.27
3(1.5, inf]3003100.00%28271765.86%16.06
合计380803493031508.27%11.09
2026-11-02T16:28:09.683140 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调下降(8.79% → 5.86%),WOE 单调上升(-0.0667 → 0.3705),符合风控分箱预期。IV=0.0203,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 2 箱 WOE<0(风险高于平均)

第 1 箱((-inf, 0.5])为坏率最高区间:坏率高达 8.79%,可作为拒绝或重点风控策略的阈值参考

setl_rem_prcp_paym_last_month_lo
组号ValueTot. Records累计占比#Goods#BadsBad RateG/B Odds
1(-inf, -4999.0]28997.62%27571424.90%19.42
2(-4999.0, inf]35161100.00%3215430078.55%10.69
合计380603491131498.27%11.09
2026-11-02T16:28:09.799953 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

关键结论

单调性良好:随分栏增大,Bad Rate 单调上升(4.90% → 8.55%),WOE 单调下降(0.5604 → -0.0361),符合风控分箱预期。IV=0.0202,该变量区分度偏弱但可用。

WOE 在第 1、2 箱之间穿越 0 线:前 1 箱 WOE>0(客户质量优于整体),后 1 箱 WOE<0(风险高于平均)

第 2 箱((-4999.0, inf])为坏率最高区间:坏率高达 8.55%,可作为拒绝或重点风控策略的阈值参考

13 压力测试

注:验证压力情境下模型的有效性和区分能力稳定性:a.坏浓度更高样本 b.时间外样本 c.特定特征客群 d.极端时间样本。当前流程自动生成情景1(贡献度最高的变量调用失败,20261026 批注2)与情景2(客群资质变差——跨时间样本剔除一半好样本,批注16)。

测试情景1:贡献度最高的变量调用失败

变量贡献度

变量贡献度
t_ps_od_days_max100.0000
available_limit_amt77.0000
cre_l90_used_limit_amt_min75.0000
f_ps_od_days_max69.0000
shd_pre_repay_amt_rate63.0000
apprv_num_m159.0000
ps_int_amt_sum47.0000
cre_l60_used_limit_rate_min44.0000
s_ps_od_days_max43.0000
min_apply_amt42.0000
失效变量贡献度口径值KS(前 → 后)AUC(前 → 后)KS 衰减AUC 衰减
f_ps_od_days_max13186.70190.3040 → 0.29240.7082 → 0.6958-0.0116-0.0124

注:失效处理口径:LR 将该变量 WOE 置 0(中性箱)后按系数重算概率;树模型将该特征置缺失重打分;评估样本 = 跨时间验证样本。衰减幅度反映模型对头部变量的依赖程度。

区分度验证 — LightGBM @ 跨时间样本

KS = 0.3040,GINI = 0.4164

2026-11-02T16:28:09.937803 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/ 2026-11-02T16:28:10.072389 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

测试情景2:客群资质变差

指标数值
样本量34730 → 18581(剔除好样本 16149)
KS(前 → 后)0.3040 → 0.3092(Δ 0.0052)
AUC(前 → 后)0.7082 → 0.7099(Δ 0.0017)
GINI(后)0.4198
2026-11-02T16:28:10.208139 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/ 2026-11-02T16:28:10.422861 image/svg+xml Matplotlib v3.10.9, https://matplotlib.org/

注:情景2 模拟客群资质变差的压力情形(好客户变少、坏客户占比升高):随机剔除一半好样本(固定 seed=42,可复现)后复算 KS/AUC。