Research Article
Research on Default Prediction for Credit Card Users Based on XGBoost-LSTM Model
Table 2
Extracted features from transaction data.
| Transaction type | First month | Second month | Third month | Max | Min | Mean | Others |
| Amount of consumption | c_sum_1 | c_sum_2 | c_sum_3 | c_sum_max | c_sum_min | c_sum_mean | c_sum_max/credit | Number of consumption | c_num_1 | c_num_2 | c_num_3 | c_num_max | c_num_min | c_num_mean | | Amount of fees | f_sum_1 | f_sum_2 | f_sum_3 | f_sum_max | f_sum_min | f_sum_mean | f_sum_max/credit | Number of fees | f_num_1 | f_num_2 | f_num_3 | f_num_max | f_num_min | f_num_mean | | Amount of repayments | r_sum_1 | r_sum_2 | r_sum_3 | r_sum_max | r_sum_min | r_sum_mean | r_sum_max/credit | Number of repayments | r_num_1 | r_num_2 | r_num_3 | r_num_max | r_num_min | r_num_mean | | Amount of cash withdrawals | cw_sum_1 | cw_sum_2 | cw_sum_3 | cw_sum_max | cw_sum_min | cw_sum_mean | cw_sum_max/credit | Number of cash withdrawals | cw_num_1 | cw_num_2 | cw_num_3 | cw_num_max | cw_num_min | cw_num_mean | | Amount of penalty | penalty_sum_1 | penalty_sum_2 | penalty_sum_3 | penalty_sum_max | penalty_sum_min | penalty_sum_median | penalty_sum_max/credit | Number of penalty | penalty_num_1 | penalty_num_2 | penalty_num_3 | penalty_num_max | penalty_num_min | penalty_num | | Repayment days in advance | pre_pay_days_1 | pre_pay_days_2 | pre_pay_days_3 | pre_pay_days_max | pre_pay_days_min | pre_pay_days_median | | Number of transaction flow | event_counts_1 | event_counts_2 | event_counts_3 | event_counts_max | event_counts_min | event_counts_median | event_counts_sum | Main transaction types | trans_type_mode_1 | trans_type_mode_2 | trans_type_mode_3 | | | | trans_type_mode_all |
|
|