Research Article
Research on Default Prediction for Credit Card Users Based on XGBoost-LSTM Model
Table 1
Example of transactions flow data.
| User ID | Input date | Input time | Transaction type | Bill amount | Credit limit |
| 100001 | 20170605 | 15:25 | 4110 | 48.28 | 50000 | 100001 | 20170605 | 15:25 | 2110 | 5000.00 | 50000 | 100001 | 20170604 | 10:51 | 7098 | −10500.00 | 50000 | 100001 | 20170605 | 15:26 | 4110 | 50.00 | 50000 | 100001 | 20170610 | 23:58 | 5000 | 148.09 | 50000 | 100001 | 20170605 | 15:26 | 2110 | 5000.00 | 50000 | 100002 | 20170605 | 21:05 | 7038 | −600.00 | 100000 | 100002 | 20170610 | 20:04 | 3254 | 1116.00 | 100000 | 100003 | 20170610 | 20:04 | 3254 | 1813.50 | 20000 | 100004 | 20170605 | 14:01 | 2000 | 6600.00 | 50000 | 100004 | 20170610 | 20:04 | 3250 | 80.00 | 50000 | 100004 | 20170610 | 20:04 | 3250 | 25.60 | 50000 | … | … | … | … | … | … |
|
|