edit feature selection
Browse files- Model Building.ipynb +297 -29
Model Building.ipynb
CHANGED
@@ -1068,19 +1068,287 @@
|
|
1068 |
},
|
1069 |
{
|
1070 |
"cell_type": "code",
|
1071 |
-
"execution_count":
|
1072 |
"id": "d66743ec",
|
1073 |
"metadata": {},
|
1074 |
-
"outputs": [
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1075 |
"source": [
|
1076 |
"from feature_engine.selection import DropCorrelatedFeatures\n",
|
1077 |
"\n",
|
1078 |
-
"fes = DropCorrelatedFeatures(threshold=0.
|
|
|
|
|
1079 |
]
|
1080 |
},
|
1081 |
{
|
1082 |
"cell_type": "code",
|
1083 |
-
"execution_count":
|
1084 |
"id": "4797c378",
|
1085 |
"metadata": {},
|
1086 |
"outputs": [],
|
@@ -1096,7 +1364,7 @@
|
|
1096 |
},
|
1097 |
{
|
1098 |
"cell_type": "code",
|
1099 |
-
"execution_count":
|
1100 |
"id": "6c8a6fd9",
|
1101 |
"metadata": {},
|
1102 |
"outputs": [],
|
@@ -1137,7 +1405,7 @@
|
|
1137 |
},
|
1138 |
{
|
1139 |
"cell_type": "code",
|
1140 |
-
"execution_count":
|
1141 |
"id": "077ca8bd",
|
1142 |
"metadata": {},
|
1143 |
"outputs": [
|
@@ -1151,11 +1419,11 @@
|
|
1151 |
"Average precision: 0.606 (+/- 0.005)\n",
|
1152 |
"F1 score: 0.596 (+/- 0.005)\n",
|
1153 |
"Balanced accuracy: 0.754 (+/- 0.004)\n",
|
1154 |
-
"Average precision: 0.
|
1155 |
-
"F1 score: 0.
|
1156 |
-
"Balanced accuracy: 0.
|
1157 |
-
"Average precision: 0.
|
1158 |
-
"F1 score: 0.596 (+/- 0.
|
1159 |
"Balanced accuracy: 0.754 (+/- 0.004)\n",
|
1160 |
"Average precision: 0.627 (+/- 0.004)\n",
|
1161 |
"F1 score: 0.608 (+/- 0.006)\n",
|
@@ -1238,7 +1506,7 @@
|
|
1238 |
},
|
1239 |
{
|
1240 |
"cell_type": "code",
|
1241 |
-
"execution_count":
|
1242 |
"id": "e4c87875",
|
1243 |
"metadata": {},
|
1244 |
"outputs": [
|
@@ -1301,21 +1569,21 @@
|
|
1301 |
" </tr>\n",
|
1302 |
" <tr>\n",
|
1303 |
" <th>both</th>\n",
|
1304 |
-
" <td>0.
|
1305 |
-
" <td>0.
|
1306 |
-
" <td>0.
|
1307 |
-
" <td>0.
|
1308 |
-
" <td>0.
|
1309 |
-
" <td>0.
|
1310 |
" </tr>\n",
|
1311 |
" <tr>\n",
|
1312 |
" <th>feature_selection</th>\n",
|
1313 |
-
" <td>0.
|
1314 |
-
" <td>0.
|
1315 |
-
" <td>0.
|
1316 |
-
" <td>0.
|
1317 |
-
" <td>0.
|
1318 |
-
" <td>0.
|
1319 |
" </tr>\n",
|
1320 |
" <tr>\n",
|
1321 |
" <th>oversampling</th>\n",
|
@@ -1344,8 +1612,8 @@
|
|
1344 |
"score average_precision balanced_accuracy f1_score \n",
|
1345 |
"model \n",
|
1346 |
"baseline 0.630397 0.687034 0.532710 \n",
|
1347 |
-
"both 0.
|
1348 |
-
"feature_selection 0.
|
1349 |
"oversampling 0.605969 0.754218 0.596470 \n",
|
1350 |
"oversampling2 0.627238 0.753277 0.608240 \n",
|
1351 |
"\n",
|
@@ -1353,13 +1621,13 @@
|
|
1353 |
"score average_precision balanced_accuracy f1_score \n",
|
1354 |
"model \n",
|
1355 |
"baseline 0.005217 0.004777 0.008199 \n",
|
1356 |
-
"both 0.
|
1357 |
-
"feature_selection 0.
|
1358 |
"oversampling 0.005173 0.003581 0.004911 \n",
|
1359 |
"oversampling2 0.004290 0.004862 0.006231 "
|
1360 |
]
|
1361 |
},
|
1362 |
-
"execution_count":
|
1363 |
"metadata": {},
|
1364 |
"output_type": "execute_result"
|
1365 |
}
|
|
|
1068 |
},
|
1069 |
{
|
1070 |
"cell_type": "code",
|
1071 |
+
"execution_count": 22,
|
1072 |
"id": "d66743ec",
|
1073 |
"metadata": {},
|
1074 |
+
"outputs": [
|
1075 |
+
{
|
1076 |
+
"data": {
|
1077 |
+
"text/html": [
|
1078 |
+
"<div>\n",
|
1079 |
+
"<style scoped>\n",
|
1080 |
+
" .dataframe tbody tr th:only-of-type {\n",
|
1081 |
+
" vertical-align: middle;\n",
|
1082 |
+
" }\n",
|
1083 |
+
"\n",
|
1084 |
+
" .dataframe tbody tr th {\n",
|
1085 |
+
" vertical-align: top;\n",
|
1086 |
+
" }\n",
|
1087 |
+
"\n",
|
1088 |
+
" .dataframe thead th {\n",
|
1089 |
+
" text-align: right;\n",
|
1090 |
+
" }\n",
|
1091 |
+
"</style>\n",
|
1092 |
+
"<table border=\"1\" class=\"dataframe\">\n",
|
1093 |
+
" <thead>\n",
|
1094 |
+
" <tr style=\"text-align: right;\">\n",
|
1095 |
+
" <th></th>\n",
|
1096 |
+
" <th>duration_ms</th>\n",
|
1097 |
+
" <th>explicit</th>\n",
|
1098 |
+
" <th>danceability</th>\n",
|
1099 |
+
" <th>energy</th>\n",
|
1100 |
+
" <th>key</th>\n",
|
1101 |
+
" <th>mode</th>\n",
|
1102 |
+
" <th>speechiness</th>\n",
|
1103 |
+
" <th>instrumentalness</th>\n",
|
1104 |
+
" <th>liveness</th>\n",
|
1105 |
+
" <th>valence</th>\n",
|
1106 |
+
" <th>tempo</th>\n",
|
1107 |
+
" <th>time_signature</th>\n",
|
1108 |
+
" <th>track_genre</th>\n",
|
1109 |
+
" </tr>\n",
|
1110 |
+
" </thead>\n",
|
1111 |
+
" <tbody>\n",
|
1112 |
+
" <tr>\n",
|
1113 |
+
" <th>22832</th>\n",
|
1114 |
+
" <td>197424</td>\n",
|
1115 |
+
" <td>0</td>\n",
|
1116 |
+
" <td>0.247</td>\n",
|
1117 |
+
" <td>0.9370</td>\n",
|
1118 |
+
" <td>9</td>\n",
|
1119 |
+
" <td>0</td>\n",
|
1120 |
+
" <td>0.0914</td>\n",
|
1121 |
+
" <td>0.789000</td>\n",
|
1122 |
+
" <td>0.7210</td>\n",
|
1123 |
+
" <td>0.1350</td>\n",
|
1124 |
+
" <td>175.189</td>\n",
|
1125 |
+
" <td>4</td>\n",
|
1126 |
+
" <td>drum-and-bass</td>\n",
|
1127 |
+
" </tr>\n",
|
1128 |
+
" <tr>\n",
|
1129 |
+
" <th>79002</th>\n",
|
1130 |
+
" <td>84948</td>\n",
|
1131 |
+
" <td>0</td>\n",
|
1132 |
+
" <td>0.282</td>\n",
|
1133 |
+
" <td>0.0663</td>\n",
|
1134 |
+
" <td>0</td>\n",
|
1135 |
+
" <td>0</td>\n",
|
1136 |
+
" <td>0.0448</td>\n",
|
1137 |
+
" <td>0.930000</td>\n",
|
1138 |
+
" <td>0.0985</td>\n",
|
1139 |
+
" <td>0.1550</td>\n",
|
1140 |
+
" <td>209.557</td>\n",
|
1141 |
+
" <td>4</td>\n",
|
1142 |
+
" <td>turkish</td>\n",
|
1143 |
+
" </tr>\n",
|
1144 |
+
" <tr>\n",
|
1145 |
+
" <th>27606</th>\n",
|
1146 |
+
" <td>284307</td>\n",
|
1147 |
+
" <td>0</td>\n",
|
1148 |
+
" <td>0.492</td>\n",
|
1149 |
+
" <td>0.4740</td>\n",
|
1150 |
+
" <td>0</td>\n",
|
1151 |
+
" <td>1</td>\n",
|
1152 |
+
" <td>0.0351</td>\n",
|
1153 |
+
" <td>0.000022</td>\n",
|
1154 |
+
" <td>0.0855</td>\n",
|
1155 |
+
" <td>0.2700</td>\n",
|
1156 |
+
" <td>130.143</td>\n",
|
1157 |
+
" <td>4</td>\n",
|
1158 |
+
" <td>folk</td>\n",
|
1159 |
+
" </tr>\n",
|
1160 |
+
" <tr>\n",
|
1161 |
+
" <th>80020</th>\n",
|
1162 |
+
" <td>292693</td>\n",
|
1163 |
+
" <td>0</td>\n",
|
1164 |
+
" <td>0.327</td>\n",
|
1165 |
+
" <td>0.2690</td>\n",
|
1166 |
+
" <td>4</td>\n",
|
1167 |
+
" <td>1</td>\n",
|
1168 |
+
" <td>0.0339</td>\n",
|
1169 |
+
" <td>0.000234</td>\n",
|
1170 |
+
" <td>0.1080</td>\n",
|
1171 |
+
" <td>0.0857</td>\n",
|
1172 |
+
" <td>141.514</td>\n",
|
1173 |
+
" <td>4</td>\n",
|
1174 |
+
" <td>world-music</td>\n",
|
1175 |
+
" </tr>\n",
|
1176 |
+
" <tr>\n",
|
1177 |
+
" <th>49203</th>\n",
|
1178 |
+
" <td>320960</td>\n",
|
1179 |
+
" <td>0</td>\n",
|
1180 |
+
" <td>0.822</td>\n",
|
1181 |
+
" <td>0.4460</td>\n",
|
1182 |
+
" <td>7</td>\n",
|
1183 |
+
" <td>0</td>\n",
|
1184 |
+
" <td>0.0321</td>\n",
|
1185 |
+
" <td>0.001020</td>\n",
|
1186 |
+
" <td>0.1470</td>\n",
|
1187 |
+
" <td>0.6540</td>\n",
|
1188 |
+
" <td>119.994</td>\n",
|
1189 |
+
" <td>4</td>\n",
|
1190 |
+
" <td>k-pop</td>\n",
|
1191 |
+
" </tr>\n",
|
1192 |
+
" <tr>\n",
|
1193 |
+
" <th>...</th>\n",
|
1194 |
+
" <td>...</td>\n",
|
1195 |
+
" <td>...</td>\n",
|
1196 |
+
" <td>...</td>\n",
|
1197 |
+
" <td>...</td>\n",
|
1198 |
+
" <td>...</td>\n",
|
1199 |
+
" <td>...</td>\n",
|
1200 |
+
" <td>...</td>\n",
|
1201 |
+
" <td>...</td>\n",
|
1202 |
+
" <td>...</td>\n",
|
1203 |
+
" <td>...</td>\n",
|
1204 |
+
" <td>...</td>\n",
|
1205 |
+
" <td>...</td>\n",
|
1206 |
+
" <td>...</td>\n",
|
1207 |
+
" </tr>\n",
|
1208 |
+
" <tr>\n",
|
1209 |
+
" <th>11265</th>\n",
|
1210 |
+
" <td>454988</td>\n",
|
1211 |
+
" <td>0</td>\n",
|
1212 |
+
" <td>0.744</td>\n",
|
1213 |
+
" <td>0.8100</td>\n",
|
1214 |
+
" <td>1</td>\n",
|
1215 |
+
" <td>1</td>\n",
|
1216 |
+
" <td>0.0473</td>\n",
|
1217 |
+
" <td>0.907000</td>\n",
|
1218 |
+
" <td>0.0606</td>\n",
|
1219 |
+
" <td>0.7150</td>\n",
|
1220 |
+
" <td>123.018</td>\n",
|
1221 |
+
" <td>4</td>\n",
|
1222 |
+
" <td>chicago-house</td>\n",
|
1223 |
+
" </tr>\n",
|
1224 |
+
" <tr>\n",
|
1225 |
+
" <th>51104</th>\n",
|
1226 |
+
" <td>210367</td>\n",
|
1227 |
+
" <td>0</td>\n",
|
1228 |
+
" <td>0.817</td>\n",
|
1229 |
+
" <td>0.5390</td>\n",
|
1230 |
+
" <td>6</td>\n",
|
1231 |
+
" <td>0</td>\n",
|
1232 |
+
" <td>0.0621</td>\n",
|
1233 |
+
" <td>0.000496</td>\n",
|
1234 |
+
" <td>0.0990</td>\n",
|
1235 |
+
" <td>0.1580</td>\n",
|
1236 |
+
" <td>97.062</td>\n",
|
1237 |
+
" <td>4</td>\n",
|
1238 |
+
" <td>latin</td>\n",
|
1239 |
+
" </tr>\n",
|
1240 |
+
" <tr>\n",
|
1241 |
+
" <th>65722</th>\n",
|
1242 |
+
" <td>152093</td>\n",
|
1243 |
+
" <td>0</td>\n",
|
1244 |
+
" <td>0.545</td>\n",
|
1245 |
+
" <td>0.3350</td>\n",
|
1246 |
+
" <td>10</td>\n",
|
1247 |
+
" <td>1</td>\n",
|
1248 |
+
" <td>0.0300</td>\n",
|
1249 |
+
" <td>0.000714</td>\n",
|
1250 |
+
" <td>0.1770</td>\n",
|
1251 |
+
" <td>0.3060</td>\n",
|
1252 |
+
" <td>119.617</td>\n",
|
1253 |
+
" <td>4</td>\n",
|
1254 |
+
" <td>rockabilly</td>\n",
|
1255 |
+
" </tr>\n",
|
1256 |
+
" <tr>\n",
|
1257 |
+
" <th>5218</th>\n",
|
1258 |
+
" <td>128600</td>\n",
|
1259 |
+
" <td>0</td>\n",
|
1260 |
+
" <td>0.504</td>\n",
|
1261 |
+
" <td>0.5770</td>\n",
|
1262 |
+
" <td>7</td>\n",
|
1263 |
+
" <td>0</td>\n",
|
1264 |
+
" <td>0.0315</td>\n",
|
1265 |
+
" <td>0.215000</td>\n",
|
1266 |
+
" <td>0.1070</td>\n",
|
1267 |
+
" <td>0.7210</td>\n",
|
1268 |
+
" <td>145.168</td>\n",
|
1269 |
+
" <td>3</td>\n",
|
1270 |
+
" <td>black-metal</td>\n",
|
1271 |
+
" </tr>\n",
|
1272 |
+
" <tr>\n",
|
1273 |
+
" <th>41260</th>\n",
|
1274 |
+
" <td>184370</td>\n",
|
1275 |
+
" <td>0</td>\n",
|
1276 |
+
" <td>0.461</td>\n",
|
1277 |
+
" <td>0.5000</td>\n",
|
1278 |
+
" <td>1</td>\n",
|
1279 |
+
" <td>1</td>\n",
|
1280 |
+
" <td>0.0614</td>\n",
|
1281 |
+
" <td>0.030000</td>\n",
|
1282 |
+
" <td>0.1160</td>\n",
|
1283 |
+
" <td>0.5720</td>\n",
|
1284 |
+
" <td>76.176</td>\n",
|
1285 |
+
" <td>4</td>\n",
|
1286 |
+
" <td>hip-hop</td>\n",
|
1287 |
+
" </tr>\n",
|
1288 |
+
" </tbody>\n",
|
1289 |
+
"</table>\n",
|
1290 |
+
"<p>64334 rows × 13 columns</p>\n",
|
1291 |
+
"</div>"
|
1292 |
+
],
|
1293 |
+
"text/plain": [
|
1294 |
+
" duration_ms explicit danceability energy key mode speechiness \\\n",
|
1295 |
+
"22832 197424 0 0.247 0.9370 9 0 0.0914 \n",
|
1296 |
+
"79002 84948 0 0.282 0.0663 0 0 0.0448 \n",
|
1297 |
+
"27606 284307 0 0.492 0.4740 0 1 0.0351 \n",
|
1298 |
+
"80020 292693 0 0.327 0.2690 4 1 0.0339 \n",
|
1299 |
+
"49203 320960 0 0.822 0.4460 7 0 0.0321 \n",
|
1300 |
+
"... ... ... ... ... .. ... ... \n",
|
1301 |
+
"11265 454988 0 0.744 0.8100 1 1 0.0473 \n",
|
1302 |
+
"51104 210367 0 0.817 0.5390 6 0 0.0621 \n",
|
1303 |
+
"65722 152093 0 0.545 0.3350 10 1 0.0300 \n",
|
1304 |
+
"5218 128600 0 0.504 0.5770 7 0 0.0315 \n",
|
1305 |
+
"41260 184370 0 0.461 0.5000 1 1 0.0614 \n",
|
1306 |
+
"\n",
|
1307 |
+
" instrumentalness liveness valence tempo time_signature \\\n",
|
1308 |
+
"22832 0.789000 0.7210 0.1350 175.189 4 \n",
|
1309 |
+
"79002 0.930000 0.0985 0.1550 209.557 4 \n",
|
1310 |
+
"27606 0.000022 0.0855 0.2700 130.143 4 \n",
|
1311 |
+
"80020 0.000234 0.1080 0.0857 141.514 4 \n",
|
1312 |
+
"49203 0.001020 0.1470 0.6540 119.994 4 \n",
|
1313 |
+
"... ... ... ... ... ... \n",
|
1314 |
+
"11265 0.907000 0.0606 0.7150 123.018 4 \n",
|
1315 |
+
"51104 0.000496 0.0990 0.1580 97.062 4 \n",
|
1316 |
+
"65722 0.000714 0.1770 0.3060 119.617 4 \n",
|
1317 |
+
"5218 0.215000 0.1070 0.7210 145.168 3 \n",
|
1318 |
+
"41260 0.030000 0.1160 0.5720 76.176 4 \n",
|
1319 |
+
"\n",
|
1320 |
+
" track_genre \n",
|
1321 |
+
"22832 drum-and-bass \n",
|
1322 |
+
"79002 turkish \n",
|
1323 |
+
"27606 folk \n",
|
1324 |
+
"80020 world-music \n",
|
1325 |
+
"49203 k-pop \n",
|
1326 |
+
"... ... \n",
|
1327 |
+
"11265 chicago-house \n",
|
1328 |
+
"51104 latin \n",
|
1329 |
+
"65722 rockabilly \n",
|
1330 |
+
"5218 black-metal \n",
|
1331 |
+
"41260 hip-hop \n",
|
1332 |
+
"\n",
|
1333 |
+
"[64334 rows x 13 columns]"
|
1334 |
+
]
|
1335 |
+
},
|
1336 |
+
"execution_count": 22,
|
1337 |
+
"metadata": {},
|
1338 |
+
"output_type": "execute_result"
|
1339 |
+
}
|
1340 |
+
],
|
1341 |
"source": [
|
1342 |
"from feature_engine.selection import DropCorrelatedFeatures\n",
|
1343 |
"\n",
|
1344 |
+
"fes = DropCorrelatedFeatures(threshold=0.6)\n",
|
1345 |
+
"te = fes.fit_transform(X_train,y_train)\n",
|
1346 |
+
"te"
|
1347 |
]
|
1348 |
},
|
1349 |
{
|
1350 |
"cell_type": "code",
|
1351 |
+
"execution_count": 17,
|
1352 |
"id": "4797c378",
|
1353 |
"metadata": {},
|
1354 |
"outputs": [],
|
|
|
1364 |
},
|
1365 |
{
|
1366 |
"cell_type": "code",
|
1367 |
+
"execution_count": 18,
|
1368 |
"id": "6c8a6fd9",
|
1369 |
"metadata": {},
|
1370 |
"outputs": [],
|
|
|
1405 |
},
|
1406 |
{
|
1407 |
"cell_type": "code",
|
1408 |
+
"execution_count": 19,
|
1409 |
"id": "077ca8bd",
|
1410 |
"metadata": {},
|
1411 |
"outputs": [
|
|
|
1419 |
"Average precision: 0.606 (+/- 0.005)\n",
|
1420 |
"F1 score: 0.596 (+/- 0.005)\n",
|
1421 |
"Balanced accuracy: 0.754 (+/- 0.004)\n",
|
1422 |
+
"Average precision: 0.630 (+/- 0.005)\n",
|
1423 |
+
"F1 score: 0.532 (+/- 0.007)\n",
|
1424 |
+
"Balanced accuracy: 0.687 (+/- 0.004)\n",
|
1425 |
+
"Average precision: 0.606 (+/- 0.005)\n",
|
1426 |
+
"F1 score: 0.596 (+/- 0.006)\n",
|
1427 |
"Balanced accuracy: 0.754 (+/- 0.004)\n",
|
1428 |
"Average precision: 0.627 (+/- 0.004)\n",
|
1429 |
"F1 score: 0.608 (+/- 0.006)\n",
|
|
|
1506 |
},
|
1507 |
{
|
1508 |
"cell_type": "code",
|
1509 |
+
"execution_count": 20,
|
1510 |
"id": "e4c87875",
|
1511 |
"metadata": {},
|
1512 |
"outputs": [
|
|
|
1569 |
" </tr>\n",
|
1570 |
" <tr>\n",
|
1571 |
" <th>both</th>\n",
|
1572 |
+
" <td>0.605690</td>\n",
|
1573 |
+
" <td>0.754049</td>\n",
|
1574 |
+
" <td>0.596330</td>\n",
|
1575 |
+
" <td>0.005015</td>\n",
|
1576 |
+
" <td>0.004315</td>\n",
|
1577 |
+
" <td>0.005606</td>\n",
|
1578 |
" </tr>\n",
|
1579 |
" <tr>\n",
|
1580 |
" <th>feature_selection</th>\n",
|
1581 |
+
" <td>0.630199</td>\n",
|
1582 |
+
" <td>0.686720</td>\n",
|
1583 |
+
" <td>0.532165</td>\n",
|
1584 |
+
" <td>0.005109</td>\n",
|
1585 |
+
" <td>0.004134</td>\n",
|
1586 |
+
" <td>0.007051</td>\n",
|
1587 |
" </tr>\n",
|
1588 |
" <tr>\n",
|
1589 |
" <th>oversampling</th>\n",
|
|
|
1612 |
"score average_precision balanced_accuracy f1_score \n",
|
1613 |
"model \n",
|
1614 |
"baseline 0.630397 0.687034 0.532710 \n",
|
1615 |
+
"both 0.605690 0.754049 0.596330 \n",
|
1616 |
+
"feature_selection 0.630199 0.686720 0.532165 \n",
|
1617 |
"oversampling 0.605969 0.754218 0.596470 \n",
|
1618 |
"oversampling2 0.627238 0.753277 0.608240 \n",
|
1619 |
"\n",
|
|
|
1621 |
"score average_precision balanced_accuracy f1_score \n",
|
1622 |
"model \n",
|
1623 |
"baseline 0.005217 0.004777 0.008199 \n",
|
1624 |
+
"both 0.005015 0.004315 0.005606 \n",
|
1625 |
+
"feature_selection 0.005109 0.004134 0.007051 \n",
|
1626 |
"oversampling 0.005173 0.003581 0.004911 \n",
|
1627 |
"oversampling2 0.004290 0.004862 0.006231 "
|
1628 |
]
|
1629 |
},
|
1630 |
+
"execution_count": 20,
|
1631 |
"metadata": {},
|
1632 |
"output_type": "execute_result"
|
1633 |
}
|