upload lightgbm_splitter
Browse files- lightgbm_splitter/doc_start/metadata.json +43 -0
- lightgbm_splitter/doc_start/model.joblib +3 -0
- lightgbm_splitter/doc_start/model.txt +0 -0
- lightgbm_splitter/doc_start/report.json +318 -0
- lightgbm_splitter/signer_page/metadata.json +41 -0
- lightgbm_splitter/signer_page/model.joblib +3 -0
- lightgbm_splitter/signer_page/model.txt +0 -0
- lightgbm_splitter/signer_page/report.json +312 -0
lightgbm_splitter/doc_start/metadata.json
ADDED
|
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"task": "doc_start",
|
| 3 |
+
"features": [
|
| 4 |
+
"regime_score",
|
| 5 |
+
"org_header_score",
|
| 6 |
+
"doc_number_standalone_score",
|
| 7 |
+
"place_date_standalone_score",
|
| 8 |
+
"subject_score",
|
| 9 |
+
"doc_number_date_alignment_score",
|
| 10 |
+
"header_completeness_score",
|
| 11 |
+
"reference_line_penalty"
|
| 12 |
+
],
|
| 13 |
+
"target": "target_doc_start",
|
| 14 |
+
"threshold": 0.5099999999999999,
|
| 15 |
+
"min_recall_for_threshold": 0.985,
|
| 16 |
+
"scale_pos_weight": 3.528519855595668,
|
| 17 |
+
"train_rows": 6272,
|
| 18 |
+
"val_rows": 736,
|
| 19 |
+
"test_rows": 689,
|
| 20 |
+
"model_params": {
|
| 21 |
+
"boosting_type": "gbdt",
|
| 22 |
+
"class_weight": null,
|
| 23 |
+
"colsample_bytree": 1.0,
|
| 24 |
+
"importance_type": "split",
|
| 25 |
+
"learning_rate": 0.035,
|
| 26 |
+
"max_depth": 4,
|
| 27 |
+
"min_child_samples": 20,
|
| 28 |
+
"min_child_weight": 0.001,
|
| 29 |
+
"min_split_gain": 0.0,
|
| 30 |
+
"n_estimators": 400,
|
| 31 |
+
"n_jobs": -1,
|
| 32 |
+
"num_leaves": 15,
|
| 33 |
+
"objective": "binary",
|
| 34 |
+
"random_state": 42,
|
| 35 |
+
"reg_alpha": 0.05,
|
| 36 |
+
"reg_lambda": 0.2,
|
| 37 |
+
"subsample": 0.9,
|
| 38 |
+
"subsample_for_bin": 200000,
|
| 39 |
+
"subsample_freq": 0,
|
| 40 |
+
"scale_pos_weight": 3.528519855595668,
|
| 41 |
+
"verbosity": -1
|
| 42 |
+
}
|
| 43 |
+
}
|
lightgbm_splitter/doc_start/model.joblib
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8820e8210c7bc0431f9d4a8f99f76460fd79fb8a1a2b806cd039b553854c8882
|
| 3 |
+
size 573404
|
lightgbm_splitter/doc_start/model.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
lightgbm_splitter/doc_start/report.json
ADDED
|
@@ -0,0 +1,318 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"task": "doc_start",
|
| 3 |
+
"features": [
|
| 4 |
+
"regime_score",
|
| 5 |
+
"org_header_score",
|
| 6 |
+
"doc_number_standalone_score",
|
| 7 |
+
"place_date_standalone_score",
|
| 8 |
+
"subject_score",
|
| 9 |
+
"doc_number_date_alignment_score",
|
| 10 |
+
"header_completeness_score",
|
| 11 |
+
"reference_line_penalty"
|
| 12 |
+
],
|
| 13 |
+
"target": "target_doc_start",
|
| 14 |
+
"threshold": 0.5099999999999999,
|
| 15 |
+
"threshold_train_metrics": {
|
| 16 |
+
"threshold": 0.5099999999999999,
|
| 17 |
+
"accuracy": 0.9692283163265306,
|
| 18 |
+
"precision": 0.8875162548764629,
|
| 19 |
+
"recall": 0.9855595667870036,
|
| 20 |
+
"f1": 0.9339719466301745,
|
| 21 |
+
"tn": 4714,
|
| 22 |
+
"fp": 173,
|
| 23 |
+
"fn": 20,
|
| 24 |
+
"tp": 1365,
|
| 25 |
+
"support_negative": 4887,
|
| 26 |
+
"support_positive": 1385,
|
| 27 |
+
"roc_auc": 0.9946593740558277,
|
| 28 |
+
"average_precision": 0.9779014157905578
|
| 29 |
+
},
|
| 30 |
+
"splits": {
|
| 31 |
+
"train": {
|
| 32 |
+
"threshold": 0.5099999999999999,
|
| 33 |
+
"accuracy": 0.9692283163265306,
|
| 34 |
+
"precision": 0.8875162548764629,
|
| 35 |
+
"recall": 0.9855595667870036,
|
| 36 |
+
"f1": 0.9339719466301745,
|
| 37 |
+
"tn": 4714,
|
| 38 |
+
"fp": 173,
|
| 39 |
+
"fn": 20,
|
| 40 |
+
"tp": 1365,
|
| 41 |
+
"support_negative": 4887,
|
| 42 |
+
"support_positive": 1385,
|
| 43 |
+
"roc_auc": 0.9946593740558277,
|
| 44 |
+
"average_precision": 0.9779014157905578,
|
| 45 |
+
"topk": [
|
| 46 |
+
{
|
| 47 |
+
"k": 1,
|
| 48 |
+
"docs_with_positive": 1384,
|
| 49 |
+
"hit": 1363,
|
| 50 |
+
"recall": 0.9848265895953757,
|
| 51 |
+
"exact_top1": 0.9841040462427746
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"k": 2,
|
| 55 |
+
"docs_with_positive": 1384,
|
| 56 |
+
"hit": 1378,
|
| 57 |
+
"recall": 0.9956647398843931,
|
| 58 |
+
"exact_top1": null
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"k": 3,
|
| 62 |
+
"docs_with_positive": 1384,
|
| 63 |
+
"hit": 1380,
|
| 64 |
+
"recall": 0.9971098265895953,
|
| 65 |
+
"exact_top1": null
|
| 66 |
+
}
|
| 67 |
+
]
|
| 68 |
+
},
|
| 69 |
+
"val": {
|
| 70 |
+
"threshold": 0.5099999999999999,
|
| 71 |
+
"accuracy": 0.9605978260869565,
|
| 72 |
+
"precision": 0.9052631578947369,
|
| 73 |
+
"recall": 0.9398907103825137,
|
| 74 |
+
"f1": 0.9222520107238605,
|
| 75 |
+
"tn": 535,
|
| 76 |
+
"fp": 18,
|
| 77 |
+
"fn": 11,
|
| 78 |
+
"tp": 172,
|
| 79 |
+
"support_negative": 553,
|
| 80 |
+
"support_positive": 183,
|
| 81 |
+
"roc_auc": 0.9843526121799623,
|
| 82 |
+
"average_precision": 0.9526122962376756,
|
| 83 |
+
"topk": [
|
| 84 |
+
{
|
| 85 |
+
"k": 1,
|
| 86 |
+
"docs_with_positive": 182,
|
| 87 |
+
"hit": 177,
|
| 88 |
+
"recall": 0.9725274725274725,
|
| 89 |
+
"exact_top1": 0.967032967032967
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"k": 2,
|
| 93 |
+
"docs_with_positive": 182,
|
| 94 |
+
"hit": 180,
|
| 95 |
+
"recall": 0.989010989010989,
|
| 96 |
+
"exact_top1": null
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"k": 3,
|
| 100 |
+
"docs_with_positive": 182,
|
| 101 |
+
"hit": 181,
|
| 102 |
+
"recall": 0.9945054945054945,
|
| 103 |
+
"exact_top1": null
|
| 104 |
+
}
|
| 105 |
+
]
|
| 106 |
+
},
|
| 107 |
+
"test": {
|
| 108 |
+
"threshold": 0.5099999999999999,
|
| 109 |
+
"accuracy": 0.9724238026124818,
|
| 110 |
+
"precision": 0.9090909090909091,
|
| 111 |
+
"recall": 0.9815950920245399,
|
| 112 |
+
"f1": 0.943952802359882,
|
| 113 |
+
"tn": 510,
|
| 114 |
+
"fp": 16,
|
| 115 |
+
"fn": 3,
|
| 116 |
+
"tp": 160,
|
| 117 |
+
"support_negative": 526,
|
| 118 |
+
"support_positive": 163,
|
| 119 |
+
"roc_auc": 0.9902551960624228,
|
| 120 |
+
"average_precision": 0.9692922817010611,
|
| 121 |
+
"topk": [
|
| 122 |
+
{
|
| 123 |
+
"k": 1,
|
| 124 |
+
"docs_with_positive": 163,
|
| 125 |
+
"hit": 159,
|
| 126 |
+
"recall": 0.9754601226993865,
|
| 127 |
+
"exact_top1": 0.9754601226993865
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"k": 2,
|
| 131 |
+
"docs_with_positive": 163,
|
| 132 |
+
"hit": 161,
|
| 133 |
+
"recall": 0.9877300613496932,
|
| 134 |
+
"exact_top1": null
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"k": 3,
|
| 138 |
+
"docs_with_positive": 163,
|
| 139 |
+
"hit": 161,
|
| 140 |
+
"recall": 0.9877300613496932,
|
| 141 |
+
"exact_top1": null
|
| 142 |
+
}
|
| 143 |
+
]
|
| 144 |
+
}
|
| 145 |
+
},
|
| 146 |
+
"feature_importance_gain": {
|
| 147 |
+
"regime_score": 1574.2763702869415,
|
| 148 |
+
"org_header_score": 5081.318770746235,
|
| 149 |
+
"doc_number_standalone_score": 4824.962701022625,
|
| 150 |
+
"place_date_standalone_score": 4106.03080637753,
|
| 151 |
+
"subject_score": 2106.368152488023,
|
| 152 |
+
"doc_number_date_alignment_score": 1393.102038219571,
|
| 153 |
+
"header_completeness_score": 142260.87477704883,
|
| 154 |
+
"reference_line_penalty": 1573.9045002609491
|
| 155 |
+
},
|
| 156 |
+
"feature_importance_split": {
|
| 157 |
+
"regime_score": 216,
|
| 158 |
+
"org_header_score": 587,
|
| 159 |
+
"doc_number_standalone_score": 661,
|
| 160 |
+
"place_date_standalone_score": 400,
|
| 161 |
+
"subject_score": 347,
|
| 162 |
+
"doc_number_date_alignment_score": 315,
|
| 163 |
+
"header_completeness_score": 1388,
|
| 164 |
+
"reference_line_penalty": 425
|
| 165 |
+
},
|
| 166 |
+
"threshold_trials_top10": [
|
| 167 |
+
{
|
| 168 |
+
"threshold": 0.7499999999999999,
|
| 169 |
+
"accuracy": 0.9754464285714286,
|
| 170 |
+
"precision": 0.9224433768016472,
|
| 171 |
+
"recall": 0.9703971119133574,
|
| 172 |
+
"f1": 0.9458128078817734,
|
| 173 |
+
"tn": 4774,
|
| 174 |
+
"fp": 113,
|
| 175 |
+
"fn": 41,
|
| 176 |
+
"tp": 1344,
|
| 177 |
+
"support_negative": 4887,
|
| 178 |
+
"support_positive": 1385,
|
| 179 |
+
"roc_auc": 0.9946593740558277,
|
| 180 |
+
"average_precision": 0.9779014157905578
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"threshold": 0.7599999999999999,
|
| 184 |
+
"accuracy": 0.9752869897959183,
|
| 185 |
+
"precision": 0.922971114167813,
|
| 186 |
+
"recall": 0.9689530685920578,
|
| 187 |
+
"f1": 0.9454033110250089,
|
| 188 |
+
"tn": 4775,
|
| 189 |
+
"fp": 112,
|
| 190 |
+
"fn": 43,
|
| 191 |
+
"tp": 1342,
|
| 192 |
+
"support_negative": 4887,
|
| 193 |
+
"support_positive": 1385,
|
| 194 |
+
"roc_auc": 0.9946593740558277,
|
| 195 |
+
"average_precision": 0.9779014157905578
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"threshold": 0.74,
|
| 199 |
+
"accuracy": 0.9751275510204082,
|
| 200 |
+
"precision": 0.9211788896504455,
|
| 201 |
+
"recall": 0.9703971119133574,
|
| 202 |
+
"f1": 0.9451476793248945,
|
| 203 |
+
"tn": 4772,
|
| 204 |
+
"fp": 115,
|
| 205 |
+
"fn": 41,
|
| 206 |
+
"tp": 1344,
|
| 207 |
+
"support_negative": 4887,
|
| 208 |
+
"support_positive": 1385,
|
| 209 |
+
"roc_auc": 0.9946593740558277,
|
| 210 |
+
"average_precision": 0.9779014157905578
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"threshold": 0.72,
|
| 214 |
+
"accuracy": 0.974968112244898,
|
| 215 |
+
"precision": 0.9188267394270123,
|
| 216 |
+
"recall": 0.9725631768953069,
|
| 217 |
+
"f1": 0.9449316029463346,
|
| 218 |
+
"tn": 4768,
|
| 219 |
+
"fp": 119,
|
| 220 |
+
"fn": 38,
|
| 221 |
+
"tp": 1347,
|
| 222 |
+
"support_negative": 4887,
|
| 223 |
+
"support_positive": 1385,
|
| 224 |
+
"roc_auc": 0.9946593740558277,
|
| 225 |
+
"average_precision": 0.9779014157905578
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"threshold": 0.7813619221894184,
|
| 229 |
+
"accuracy": 0.9751275510204082,
|
| 230 |
+
"precision": 0.9258489258489259,
|
| 231 |
+
"recall": 0.9646209386281589,
|
| 232 |
+
"f1": 0.9448373408769448,
|
| 233 |
+
"tn": 4780,
|
| 234 |
+
"fp": 107,
|
| 235 |
+
"fn": 49,
|
| 236 |
+
"tp": 1336,
|
| 237 |
+
"support_negative": 4887,
|
| 238 |
+
"support_positive": 1385,
|
| 239 |
+
"roc_auc": 0.9946593740558277,
|
| 240 |
+
"average_precision": 0.9779014157905578
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"threshold": 0.71,
|
| 244 |
+
"accuracy": 0.9748086734693877,
|
| 245 |
+
"precision": 0.918200408997955,
|
| 246 |
+
"recall": 0.9725631768953069,
|
| 247 |
+
"f1": 0.9446002805049089,
|
| 248 |
+
"tn": 4767,
|
| 249 |
+
"fp": 120,
|
| 250 |
+
"fn": 38,
|
| 251 |
+
"tp": 1347,
|
| 252 |
+
"support_negative": 4887,
|
| 253 |
+
"support_positive": 1385,
|
| 254 |
+
"roc_auc": 0.9946593740558277,
|
| 255 |
+
"average_precision": 0.9779014157905578
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"threshold": 0.73,
|
| 259 |
+
"accuracy": 0.9748086734693877,
|
| 260 |
+
"precision": 0.9193438140806562,
|
| 261 |
+
"recall": 0.9711191335740073,
|
| 262 |
+
"f1": 0.9445224719101124,
|
| 263 |
+
"tn": 4769,
|
| 264 |
+
"fp": 118,
|
| 265 |
+
"fn": 40,
|
| 266 |
+
"tp": 1345,
|
| 267 |
+
"support_negative": 4887,
|
| 268 |
+
"support_positive": 1385,
|
| 269 |
+
"roc_auc": 0.9946593740558277,
|
| 270 |
+
"average_precision": 0.9779014157905578
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"threshold": 0.7799999999999999,
|
| 274 |
+
"accuracy": 0.974968112244898,
|
| 275 |
+
"precision": 0.925207756232687,
|
| 276 |
+
"recall": 0.9646209386281589,
|
| 277 |
+
"f1": 0.944503358077059,
|
| 278 |
+
"tn": 4779,
|
| 279 |
+
"fp": 108,
|
| 280 |
+
"fn": 49,
|
| 281 |
+
"tp": 1336,
|
| 282 |
+
"support_negative": 4887,
|
| 283 |
+
"support_positive": 1385,
|
| 284 |
+
"roc_auc": 0.9946593740558277,
|
| 285 |
+
"average_precision": 0.9779014157905578
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"threshold": 0.7699999999999999,
|
| 289 |
+
"accuracy": 0.9746492346938775,
|
| 290 |
+
"precision": 0.9239280774550485,
|
| 291 |
+
"recall": 0.9646209386281589,
|
| 292 |
+
"f1": 0.9438361003179089,
|
| 293 |
+
"tn": 4777,
|
| 294 |
+
"fp": 110,
|
| 295 |
+
"fn": 49,
|
| 296 |
+
"tp": 1336,
|
| 297 |
+
"support_negative": 4887,
|
| 298 |
+
"support_positive": 1385,
|
| 299 |
+
"roc_auc": 0.9946593740558277,
|
| 300 |
+
"average_precision": 0.9779014157905578
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"threshold": 0.7899999999999999,
|
| 304 |
+
"accuracy": 0.9746492346938775,
|
| 305 |
+
"precision": 0.9268802228412256,
|
| 306 |
+
"recall": 0.9610108303249097,
|
| 307 |
+
"f1": 0.9436370081531372,
|
| 308 |
+
"tn": 4782,
|
| 309 |
+
"fp": 105,
|
| 310 |
+
"fn": 54,
|
| 311 |
+
"tp": 1331,
|
| 312 |
+
"support_negative": 4887,
|
| 313 |
+
"support_positive": 1385,
|
| 314 |
+
"roc_auc": 0.9946593740558277,
|
| 315 |
+
"average_precision": 0.9779014157905578
|
| 316 |
+
}
|
| 317 |
+
]
|
| 318 |
+
}
|
lightgbm_splitter/signer_page/metadata.json
ADDED
|
@@ -0,0 +1,41 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"task": "signer_page",
|
| 3 |
+
"features": [
|
| 4 |
+
"recipients_score",
|
| 5 |
+
"signer_role_score",
|
| 6 |
+
"signer_name_score",
|
| 7 |
+
"has_noi_nhan_regex",
|
| 8 |
+
"has_tm_kt_tl_tuq_regex",
|
| 9 |
+
"relative_page_position"
|
| 10 |
+
],
|
| 11 |
+
"target": "target_signer_page",
|
| 12 |
+
"threshold": 0.38999999999999996,
|
| 13 |
+
"min_recall_for_threshold": 0.99,
|
| 14 |
+
"scale_pos_weight": 3.0359820089955023,
|
| 15 |
+
"train_rows": 5384,
|
| 16 |
+
"val_rows": 653,
|
| 17 |
+
"test_rows": 574,
|
| 18 |
+
"model_params": {
|
| 19 |
+
"boosting_type": "gbdt",
|
| 20 |
+
"class_weight": null,
|
| 21 |
+
"colsample_bytree": 1.0,
|
| 22 |
+
"importance_type": "split",
|
| 23 |
+
"learning_rate": 0.035,
|
| 24 |
+
"max_depth": 4,
|
| 25 |
+
"min_child_samples": 20,
|
| 26 |
+
"min_child_weight": 0.001,
|
| 27 |
+
"min_split_gain": 0.0,
|
| 28 |
+
"n_estimators": 400,
|
| 29 |
+
"n_jobs": -1,
|
| 30 |
+
"num_leaves": 15,
|
| 31 |
+
"objective": "binary",
|
| 32 |
+
"random_state": 42,
|
| 33 |
+
"reg_alpha": 0.05,
|
| 34 |
+
"reg_lambda": 0.2,
|
| 35 |
+
"subsample": 0.9,
|
| 36 |
+
"subsample_for_bin": 200000,
|
| 37 |
+
"subsample_freq": 0,
|
| 38 |
+
"scale_pos_weight": 3.0359820089955023,
|
| 39 |
+
"verbosity": -1
|
| 40 |
+
}
|
| 41 |
+
}
|
lightgbm_splitter/signer_page/model.joblib
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c06f23e9b937e50655c31e9336c8387d6fc2fc7cee1e65645e2043547453c8c9
|
| 3 |
+
size 568540
|
lightgbm_splitter/signer_page/model.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
lightgbm_splitter/signer_page/report.json
ADDED
|
@@ -0,0 +1,312 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"task": "signer_page",
|
| 3 |
+
"features": [
|
| 4 |
+
"recipients_score",
|
| 5 |
+
"signer_role_score",
|
| 6 |
+
"signer_name_score",
|
| 7 |
+
"has_noi_nhan_regex",
|
| 8 |
+
"has_tm_kt_tl_tuq_regex",
|
| 9 |
+
"relative_page_position"
|
| 10 |
+
],
|
| 11 |
+
"target": "target_signer_page",
|
| 12 |
+
"threshold": 0.38999999999999996,
|
| 13 |
+
"threshold_train_metrics": {
|
| 14 |
+
"threshold": 0.38999999999999996,
|
| 15 |
+
"accuracy": 0.9853268945022289,
|
| 16 |
+
"precision": 0.9524152847873107,
|
| 17 |
+
"recall": 0.9902548725637181,
|
| 18 |
+
"f1": 0.9709665564130834,
|
| 19 |
+
"tn": 3984,
|
| 20 |
+
"fp": 66,
|
| 21 |
+
"fn": 13,
|
| 22 |
+
"tp": 1321,
|
| 23 |
+
"support_negative": 4050,
|
| 24 |
+
"support_positive": 1334,
|
| 25 |
+
"roc_auc": 0.9991101671386529,
|
| 26 |
+
"average_precision": 0.9972744880654751
|
| 27 |
+
},
|
| 28 |
+
"splits": {
|
| 29 |
+
"train": {
|
| 30 |
+
"threshold": 0.38999999999999996,
|
| 31 |
+
"accuracy": 0.9853268945022289,
|
| 32 |
+
"precision": 0.9524152847873107,
|
| 33 |
+
"recall": 0.9902548725637181,
|
| 34 |
+
"f1": 0.9709665564130834,
|
| 35 |
+
"tn": 3984,
|
| 36 |
+
"fp": 66,
|
| 37 |
+
"fn": 13,
|
| 38 |
+
"tp": 1321,
|
| 39 |
+
"support_negative": 4050,
|
| 40 |
+
"support_positive": 1334,
|
| 41 |
+
"roc_auc": 0.9991101671386529,
|
| 42 |
+
"average_precision": 0.9972744880654751,
|
| 43 |
+
"topk": [
|
| 44 |
+
{
|
| 45 |
+
"k": 1,
|
| 46 |
+
"docs_with_positive": 1334,
|
| 47 |
+
"hit": 1333,
|
| 48 |
+
"recall": 0.9992503748125937,
|
| 49 |
+
"exact_top1": 0.9992503748125937
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"k": 2,
|
| 53 |
+
"docs_with_positive": 1334,
|
| 54 |
+
"hit": 1334,
|
| 55 |
+
"recall": 1.0,
|
| 56 |
+
"exact_top1": null
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"k": 3,
|
| 60 |
+
"docs_with_positive": 1334,
|
| 61 |
+
"hit": 1334,
|
| 62 |
+
"recall": 1.0,
|
| 63 |
+
"exact_top1": null
|
| 64 |
+
}
|
| 65 |
+
]
|
| 66 |
+
},
|
| 67 |
+
"val": {
|
| 68 |
+
"threshold": 0.38999999999999996,
|
| 69 |
+
"accuracy": 0.9739663093415007,
|
| 70 |
+
"precision": 0.9193548387096774,
|
| 71 |
+
"recall": 0.9884393063583815,
|
| 72 |
+
"f1": 0.9526462395543176,
|
| 73 |
+
"tn": 465,
|
| 74 |
+
"fp": 15,
|
| 75 |
+
"fn": 2,
|
| 76 |
+
"tp": 171,
|
| 77 |
+
"support_negative": 480,
|
| 78 |
+
"support_positive": 173,
|
| 79 |
+
"roc_auc": 0.9980069845857418,
|
| 80 |
+
"average_precision": 0.994957716837136,
|
| 81 |
+
"topk": [
|
| 82 |
+
{
|
| 83 |
+
"k": 1,
|
| 84 |
+
"docs_with_positive": 173,
|
| 85 |
+
"hit": 173,
|
| 86 |
+
"recall": 1.0,
|
| 87 |
+
"exact_top1": 1.0
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"k": 2,
|
| 91 |
+
"docs_with_positive": 173,
|
| 92 |
+
"hit": 173,
|
| 93 |
+
"recall": 1.0,
|
| 94 |
+
"exact_top1": null
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"k": 3,
|
| 98 |
+
"docs_with_positive": 173,
|
| 99 |
+
"hit": 173,
|
| 100 |
+
"recall": 1.0,
|
| 101 |
+
"exact_top1": null
|
| 102 |
+
}
|
| 103 |
+
]
|
| 104 |
+
},
|
| 105 |
+
"test": {
|
| 106 |
+
"threshold": 0.38999999999999996,
|
| 107 |
+
"accuracy": 0.9843205574912892,
|
| 108 |
+
"precision": 0.9512195121951219,
|
| 109 |
+
"recall": 0.9936305732484076,
|
| 110 |
+
"f1": 0.9719626168224299,
|
| 111 |
+
"tn": 409,
|
| 112 |
+
"fp": 8,
|
| 113 |
+
"fn": 1,
|
| 114 |
+
"tp": 156,
|
| 115 |
+
"support_negative": 417,
|
| 116 |
+
"support_positive": 157,
|
| 117 |
+
"roc_auc": 0.9987474988162337,
|
| 118 |
+
"average_precision": 0.9963874949798402,
|
| 119 |
+
"topk": [
|
| 120 |
+
{
|
| 121 |
+
"k": 1,
|
| 122 |
+
"docs_with_positive": 157,
|
| 123 |
+
"hit": 157,
|
| 124 |
+
"recall": 1.0,
|
| 125 |
+
"exact_top1": 1.0
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"k": 2,
|
| 129 |
+
"docs_with_positive": 157,
|
| 130 |
+
"hit": 157,
|
| 131 |
+
"recall": 1.0,
|
| 132 |
+
"exact_top1": null
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"k": 3,
|
| 136 |
+
"docs_with_positive": 157,
|
| 137 |
+
"hit": 157,
|
| 138 |
+
"recall": 1.0,
|
| 139 |
+
"exact_top1": null
|
| 140 |
+
}
|
| 141 |
+
]
|
| 142 |
+
}
|
| 143 |
+
},
|
| 144 |
+
"feature_importance_gain": {
|
| 145 |
+
"recipients_score": 14606.816122462973,
|
| 146 |
+
"signer_role_score": 1553.787613954395,
|
| 147 |
+
"signer_name_score": 78758.13409495406,
|
| 148 |
+
"has_noi_nhan_regex": 0.0,
|
| 149 |
+
"has_tm_kt_tl_tuq_regex": 2600.9574969410896,
|
| 150 |
+
"relative_page_position": 47941.771243711
|
| 151 |
+
},
|
| 152 |
+
"feature_importance_split": {
|
| 153 |
+
"recipients_score": 474,
|
| 154 |
+
"signer_role_score": 619,
|
| 155 |
+
"signer_name_score": 1451,
|
| 156 |
+
"has_noi_nhan_regex": 0,
|
| 157 |
+
"has_tm_kt_tl_tuq_regex": 183,
|
| 158 |
+
"relative_page_position": 1542
|
| 159 |
+
},
|
| 160 |
+
"threshold_trials_top10": [
|
| 161 |
+
{
|
| 162 |
+
"threshold": 0.6699999999999999,
|
| 163 |
+
"accuracy": 0.9890416047548292,
|
| 164 |
+
"precision": 0.9718726868985936,
|
| 165 |
+
"recall": 0.9842578710644677,
|
| 166 |
+
"f1": 0.9780260707635009,
|
| 167 |
+
"tn": 4012,
|
| 168 |
+
"fp": 38,
|
| 169 |
+
"fn": 21,
|
| 170 |
+
"tp": 1313,
|
| 171 |
+
"support_negative": 4050,
|
| 172 |
+
"support_positive": 1334,
|
| 173 |
+
"roc_auc": 0.9991101671386529,
|
| 174 |
+
"average_precision": 0.9972744880654751
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"threshold": 0.6599999999999999,
|
| 178 |
+
"accuracy": 0.9888558692421991,
|
| 179 |
+
"precision": 0.9711538461538461,
|
| 180 |
+
"recall": 0.9842578710644677,
|
| 181 |
+
"f1": 0.9776619508562919,
|
| 182 |
+
"tn": 4011,
|
| 183 |
+
"fp": 39,
|
| 184 |
+
"fn": 21,
|
| 185 |
+
"tp": 1313,
|
| 186 |
+
"support_negative": 4050,
|
| 187 |
+
"support_positive": 1334,
|
| 188 |
+
"roc_auc": 0.9991101671386529,
|
| 189 |
+
"average_precision": 0.9972744880654751
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"threshold": 0.6799999999999999,
|
| 193 |
+
"accuracy": 0.9888558692421991,
|
| 194 |
+
"precision": 0.9718518518518519,
|
| 195 |
+
"recall": 0.9835082458770614,
|
| 196 |
+
"f1": 0.977645305514158,
|
| 197 |
+
"tn": 4012,
|
| 198 |
+
"fp": 38,
|
| 199 |
+
"fn": 22,
|
| 200 |
+
"tp": 1312,
|
| 201 |
+
"support_negative": 4050,
|
| 202 |
+
"support_positive": 1334,
|
| 203 |
+
"roc_auc": 0.9991101671386529,
|
| 204 |
+
"average_precision": 0.9972744880654751
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"threshold": 0.69,
|
| 208 |
+
"accuracy": 0.9888558692421991,
|
| 209 |
+
"precision": 0.9718518518518519,
|
| 210 |
+
"recall": 0.9835082458770614,
|
| 211 |
+
"f1": 0.977645305514158,
|
| 212 |
+
"tn": 4012,
|
| 213 |
+
"fp": 38,
|
| 214 |
+
"fn": 22,
|
| 215 |
+
"tp": 1312,
|
| 216 |
+
"support_negative": 4050,
|
| 217 |
+
"support_positive": 1334,
|
| 218 |
+
"roc_auc": 0.9991101671386529,
|
| 219 |
+
"average_precision": 0.9972744880654751
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"threshold": 0.7,
|
| 223 |
+
"accuracy": 0.9888558692421991,
|
| 224 |
+
"precision": 0.9718518518518519,
|
| 225 |
+
"recall": 0.9835082458770614,
|
| 226 |
+
"f1": 0.977645305514158,
|
| 227 |
+
"tn": 4012,
|
| 228 |
+
"fp": 38,
|
| 229 |
+
"fn": 22,
|
| 230 |
+
"tp": 1312,
|
| 231 |
+
"support_negative": 4050,
|
| 232 |
+
"support_positive": 1334,
|
| 233 |
+
"roc_auc": 0.9991101671386529,
|
| 234 |
+
"average_precision": 0.9972744880654751
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"threshold": 0.71,
|
| 238 |
+
"accuracy": 0.9888558692421991,
|
| 239 |
+
"precision": 0.9718518518518519,
|
| 240 |
+
"recall": 0.9835082458770614,
|
| 241 |
+
"f1": 0.977645305514158,
|
| 242 |
+
"tn": 4012,
|
| 243 |
+
"fp": 38,
|
| 244 |
+
"fn": 22,
|
| 245 |
+
"tp": 1312,
|
| 246 |
+
"support_negative": 4050,
|
| 247 |
+
"support_positive": 1334,
|
| 248 |
+
"roc_auc": 0.9991101671386529,
|
| 249 |
+
"average_precision": 0.9972744880654751
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"threshold": 0.72,
|
| 253 |
+
"accuracy": 0.9888558692421991,
|
| 254 |
+
"precision": 0.9725519287833828,
|
| 255 |
+
"recall": 0.9827586206896551,
|
| 256 |
+
"f1": 0.9776286353467561,
|
| 257 |
+
"tn": 4013,
|
| 258 |
+
"fp": 37,
|
| 259 |
+
"fn": 23,
|
| 260 |
+
"tp": 1311,
|
| 261 |
+
"support_negative": 4050,
|
| 262 |
+
"support_positive": 1334,
|
| 263 |
+
"roc_auc": 0.9991101671386529,
|
| 264 |
+
"average_precision": 0.9972744880654751
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"threshold": 0.7314328207214009,
|
| 268 |
+
"accuracy": 0.9888558692421991,
|
| 269 |
+
"precision": 0.9732540861812778,
|
| 270 |
+
"recall": 0.9820089955022488,
|
| 271 |
+
"f1": 0.9776119402985075,
|
| 272 |
+
"tn": 4014,
|
| 273 |
+
"fp": 36,
|
| 274 |
+
"fn": 24,
|
| 275 |
+
"tp": 1310,
|
| 276 |
+
"support_negative": 4050,
|
| 277 |
+
"support_positive": 1334,
|
| 278 |
+
"roc_auc": 0.9991101671386529,
|
| 279 |
+
"average_precision": 0.9972744880654751
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"threshold": 0.74,
|
| 283 |
+
"accuracy": 0.9888558692421991,
|
| 284 |
+
"precision": 0.9739583333333334,
|
| 285 |
+
"recall": 0.9812593703148426,
|
| 286 |
+
"f1": 0.977595220313667,
|
| 287 |
+
"tn": 4015,
|
| 288 |
+
"fp": 35,
|
| 289 |
+
"fn": 25,
|
| 290 |
+
"tp": 1309,
|
| 291 |
+
"support_negative": 4050,
|
| 292 |
+
"support_positive": 1334,
|
| 293 |
+
"roc_auc": 0.9991101671386529,
|
| 294 |
+
"average_precision": 0.9972744880654751
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"threshold": 0.7499999999999999,
|
| 298 |
+
"accuracy": 0.9888558692421991,
|
| 299 |
+
"precision": 0.9739583333333334,
|
| 300 |
+
"recall": 0.9812593703148426,
|
| 301 |
+
"f1": 0.977595220313667,
|
| 302 |
+
"tn": 4015,
|
| 303 |
+
"fp": 35,
|
| 304 |
+
"fn": 25,
|
| 305 |
+
"tp": 1309,
|
| 306 |
+
"support_negative": 4050,
|
| 307 |
+
"support_positive": 1334,
|
| 308 |
+
"roc_auc": 0.9991101671386529,
|
| 309 |
+
"average_precision": 0.9972744880654751
|
| 310 |
+
}
|
| 311 |
+
]
|
| 312 |
+
}
|