feature_attr_list: ['custAge', 'profession', 'marital', 'schooling', 'default', 'housing', 'loan', 'contact', 'month', 'day_of_week', 'campaign', 'pdays', 'previous', 'poutcome', 'emp.var.rate', 'cons.price.idx', 'cons.conf.idx', 'euribor3m', 'nr.employed', 'pastEmail']
target_attr_list: ['responded', 'profit']
feature_array shape: (8137, 20)
target_array shape: (8137, 2)
data_package:
feature_standard_weight_list: 58 or 54
customer_feature_standard_weight_list: 32 or 29
train_input: (7323, 58 or 54)
train_cust: (7323, 32 or 29)
train_target: (7323, 2)
val_input: (814, 58 or 54)
val_cust: (814, 32 or 29)
val_target: (814, 2)
| experiments | mode | setup | acc_most | profit_most | figure_name |
|---|---|---|---|---|---|
| baseline1 | sample | 原始的cls labels, svm | 0.917(c=0.01), 4353 | 0.82, 7731(0.343), RR=0.684 | acc_bs1_svm1.png, profit_bs1_svm1.png |
| baseline1 | sample | synthesis cls labels, svm | 0.925(c=0.01), 4353 | 0.845, 8103(0.343), RR=0.689 | acc_bs1_svm2.png, profit_bs1_svm2.png |
| baseline1 | average | 原始的cls labels, svm | 0.917(c=0.01), 4353 | 0.82, 7109(0.364) | acc_average_bs1_svm1.png, profit_average_bs1_svm1.png |
| baseline1 | average | synthesis cls labels, svm | 0.925(c=0.01), 4353 | 0.845, 8110(0.333) | acc_average_bs1_svm2.png, profit_average_bs1_svm2.png |
| experiments | mode | setup | acc_most | profit_most | figure_name |
|---|---|---|---|---|---|
| bs1 lg | sample | 原始的cls labels, svm | 0.917(c=0.01), 4353 | 0.82, 7731(0.343), RR=0.684 | acc_bs1_svm1.png, profit_bs1_svm1.png |
| bs1 lg | sample | synthesis cls labels, svm | 0.925(c=0.01), 4353 | 0.845, 8103(0.343), RR=0.689 | acc_bs1_svm2.png, profit_bs1_svm2.png |
| baseline1 | average | 原始的cls labels, svm | 0.917(c=0.01), 4353 | 0.82, 7109(0.364) | acc_average_bs1_svm1.png, profit_average_bs1_svm1.png |
| baseline1 | average | synthesis cls labels, svm | 0.925(c=0.01), 4353 | 0.845, 8110(0.333) | acc_average_bs1_svm2.png, profit_average_bs1_svm2.png |
TP: total_precision
RR: recommend_recall
| Baseline | data mode | setup | train_result | val_result |
|---|---|---|---|---|
| 1 | zero | base DT | TP:0.95 RR:0.57 Profit:62244 | TP:0.89 RR:0.36 Profit:4473 |
| 1 | average | base DT | TP:0.95 RR:0.59 Profit:64868 | TP:0.90 RR:0.33 Profit:4408 |
| 1 | sample | base DT | TP:0.94 RR:0.45 Profit:49059 | TP:0.90 RR:0.34 Profit:4839 |
| Baseline | data mode | setup | train_result | val_result |
|---|---|---|---|---|
| 1 | sample | LR, L2 norm, balance loss | TP:0.80 RR:0.74 Profit:47226 | TP:0.78 RR:0.71 Profit:5826 |
| 1 | sample | LR, L2 norm, balance data | TP:0.81 RR:0.72 Profit:47936 | TP:0.79 RR:0.68 Profit:5945 |
| 1 | sample | MLR2, L2 norm, balance loss | TP:0.81 RR:0.72 Profit:48880 | TP:0.79 RR:0.69 Profit:5946 |
| 1 | sample | MLR4, L2 norm, balance loss | TP:0.92 RR:0.99 Profit:96072 | TP:0.84 RR:0.47 Profit:4542 |