/////////////////////////////////////////// // Running convGAN-majority-full on imblearn_ozone_level /////////////////////////////////////////// Load 'data_input/imblearn_ozone_level' from imblearn Data loaded. -> Shuffling data ### Start exercise for synthetic point generator ====== Step 1/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 1/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 301, 192 GAN fn, tp: 1, 14 GAN f1 score: 0.127 GAN cohens kappa score: 0.076 -> test with 'LR' LR tn, fp: 437, 56 LR fn, tp: 1, 14 LR f1 score: 0.329 LR cohens kappa score: 0.295 LR average precision score: 0.340 -> test with 'GB' GB tn, fp: 477, 16 GB fn, tp: 9, 6 GB f1 score: 0.324 GB cohens kappa score: 0.300 -> test with 'KNN' KNN tn, fp: 386, 107 KNN fn, tp: 10, 5 KNN f1 score: 0.079 KNN cohens kappa score: 0.028 ------ Step 1/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 395, 98 GAN fn, tp: 4, 11 GAN f1 score: 0.177 GAN cohens kappa score: 0.132 -> test with 'LR' LR tn, fp: 435, 58 LR fn, tp: 4, 11 LR f1 score: 0.262 LR cohens kappa score: 0.224 LR average precision score: 0.228 -> test with 'GB' GB tn, fp: 486, 7 GB fn, tp: 9, 6 GB f1 score: 0.429 GB cohens kappa score: 0.412 -> test with 'KNN' KNN tn, fp: 423, 70 KNN fn, tp: 10, 5 KNN f1 score: 0.111 KNN cohens kappa score: 0.065 ------ Step 1/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 493, 0 GAN fn, tp: 15, 0 GAN f1 score: 0.000 GAN cohens kappa score: 0.000 -> test with 'LR' LR tn, fp: 439, 54 LR fn, tp: 4, 11 LR f1 score: 0.275 LR cohens kappa score: 0.238 LR average precision score: 0.128 -> test with 'GB' GB tn, fp: 472, 21 GB fn, tp: 10, 5 GB f1 score: 0.244 GB cohens kappa score: 0.214 -> test with 'KNN' KNN tn, fp: 390, 103 KNN fn, tp: 7, 8 KNN f1 score: 0.127 KNN cohens kappa score: 0.079 ------ Step 1/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 380, 113 GAN fn, tp: 3, 12 GAN f1 score: 0.171 GAN cohens kappa score: 0.125 -> test with 'LR' LR tn, fp: 436, 57 LR fn, tp: 6, 9 LR f1 score: 0.222 LR cohens kappa score: 0.183 LR average precision score: 0.226 -> test with 'GB' GB tn, fp: 480, 13 GB fn, tp: 9, 6 GB f1 score: 0.353 GB cohens kappa score: 0.331 -> test with 'KNN' KNN tn, fp: 407, 86 KNN fn, tp: 10, 5 KNN f1 score: 0.094 KNN cohens kappa score: 0.046 ------ Step 1/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 302, 189 GAN fn, tp: 1, 12 GAN f1 score: 0.112 GAN cohens kappa score: 0.067 -> test with 'LR' LR tn, fp: 437, 54 LR fn, tp: 3, 10 LR f1 score: 0.260 LR cohens kappa score: 0.227 LR average precision score: 0.171 -> test with 'GB' GB tn, fp: 479, 12 GB fn, tp: 11, 2 GB f1 score: 0.148 GB cohens kappa score: 0.125 -> test with 'KNN' KNN tn, fp: 395, 96 KNN fn, tp: 9, 4 KNN f1 score: 0.071 KNN cohens kappa score: 0.026 ====== Step 2/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 2/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 306, 187 GAN fn, tp: 2, 13 GAN f1 score: 0.121 GAN cohens kappa score: 0.070 -> test with 'LR' LR tn, fp: 438, 55 LR fn, tp: 6, 9 LR f1 score: 0.228 LR cohens kappa score: 0.189 LR average precision score: 0.261 -> test with 'GB' GB tn, fp: 483, 10 GB fn, tp: 12, 3 GB f1 score: 0.214 GB cohens kappa score: 0.192 -> test with 'KNN' KNN tn, fp: 406, 87 KNN fn, tp: 8, 7 KNN f1 score: 0.128 KNN cohens kappa score: 0.082 ------ Step 2/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 278, 215 GAN fn, tp: 0, 15 GAN f1 score: 0.122 GAN cohens kappa score: 0.071 -> test with 'LR' LR tn, fp: 443, 50 LR fn, tp: 6, 9 LR f1 score: 0.243 LR cohens kappa score: 0.206 LR average precision score: 0.196 -> test with 'GB' GB tn, fp: 480, 13 GB fn, tp: 9, 6 GB f1 score: 0.353 GB cohens kappa score: 0.331 -> test with 'KNN' KNN tn, fp: 402, 91 KNN fn, tp: 7, 8 KNN f1 score: 0.140 KNN cohens kappa score: 0.094 ------ Step 2/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 339, 154 GAN fn, tp: 1, 14 GAN f1 score: 0.153 GAN cohens kappa score: 0.104 -> test with 'LR' LR tn, fp: 437, 56 LR fn, tp: 1, 14 LR f1 score: 0.329 LR cohens kappa score: 0.295 LR average precision score: 0.468 -> test with 'GB' GB tn, fp: 487, 6 GB fn, tp: 9, 6 GB f1 score: 0.444 GB cohens kappa score: 0.429 -> test with 'KNN' KNN tn, fp: 390, 103 KNN fn, tp: 13, 2 KNN f1 score: 0.033 KNN cohens kappa score: -0.019 ------ Step 2/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 399, 94 GAN fn, tp: 6, 9 GAN f1 score: 0.153 GAN cohens kappa score: 0.106 -> test with 'LR' LR tn, fp: 440, 53 LR fn, tp: 5, 10 LR f1 score: 0.256 LR cohens kappa score: 0.219 LR average precision score: 0.165 -> test with 'GB' GB tn, fp: 479, 14 GB fn, tp: 9, 6 GB f1 score: 0.343 GB cohens kappa score: 0.320 -> test with 'KNN' KNN tn, fp: 415, 78 KNN fn, tp: 10, 5 KNN f1 score: 0.102 KNN cohens kappa score: 0.055 ------ Step 2/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 484, 7 GAN fn, tp: 10, 3 GAN f1 score: 0.261 GAN cohens kappa score: 0.244 -> test with 'LR' LR tn, fp: 426, 65 LR fn, tp: 3, 10 LR f1 score: 0.227 LR cohens kappa score: 0.192 LR average precision score: 0.214 -> test with 'GB' GB tn, fp: 481, 10 GB fn, tp: 9, 4 GB f1 score: 0.296 GB cohens kappa score: 0.277 -> test with 'KNN' KNN tn, fp: 402, 89 KNN fn, tp: 7, 6 KNN f1 score: 0.111 KNN cohens kappa score: 0.069 ====== Step 3/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 3/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 458, 35 GAN fn, tp: 9, 6 GAN f1 score: 0.214 GAN cohens kappa score: 0.179 -> test with 'LR' LR tn, fp: 436, 57 LR fn, tp: 3, 12 LR f1 score: 0.286 LR cohens kappa score: 0.249 LR average precision score: 0.319 -> test with 'GB' GB tn, fp: 483, 10 GB fn, tp: 12, 3 GB f1 score: 0.214 GB cohens kappa score: 0.192 -> test with 'KNN' KNN tn, fp: 411, 82 KNN fn, tp: 12, 3 KNN f1 score: 0.060 KNN cohens kappa score: 0.010 ------ Step 3/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 449, 44 GAN fn, tp: 6, 9 GAN f1 score: 0.265 GAN cohens kappa score: 0.229 -> test with 'LR' LR tn, fp: 437, 56 LR fn, tp: 5, 10 LR f1 score: 0.247 LR cohens kappa score: 0.209 LR average precision score: 0.135 -> test with 'GB' GB tn, fp: 485, 8 GB fn, tp: 9, 6 GB f1 score: 0.414 GB cohens kappa score: 0.397 -> test with 'KNN' KNN tn, fp: 411, 82 KNN fn, tp: 10, 5 KNN f1 score: 0.098 KNN cohens kappa score: 0.050 ------ Step 3/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 463, 30 GAN fn, tp: 8, 7 GAN f1 score: 0.269 GAN cohens kappa score: 0.237 -> test with 'LR' LR tn, fp: 451, 42 LR fn, tp: 5, 10 LR f1 score: 0.299 LR cohens kappa score: 0.265 LR average precision score: 0.179 -> test with 'GB' GB tn, fp: 476, 17 GB fn, tp: 7, 8 GB f1 score: 0.400 GB cohens kappa score: 0.377 -> test with 'KNN' KNN tn, fp: 393, 100 KNN fn, tp: 6, 9 KNN f1 score: 0.145 KNN cohens kappa score: 0.098 ------ Step 3/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 328, 165 GAN fn, tp: 1, 14 GAN f1 score: 0.144 GAN cohens kappa score: 0.095 -> test with 'LR' LR tn, fp: 441, 52 LR fn, tp: 5, 10 LR f1 score: 0.260 LR cohens kappa score: 0.223 LR average precision score: 0.187 -> test with 'GB' GB tn, fp: 483, 10 GB fn, tp: 8, 7 GB f1 score: 0.437 GB cohens kappa score: 0.419 -> test with 'KNN' KNN tn, fp: 407, 86 KNN fn, tp: 11, 4 KNN f1 score: 0.076 KNN cohens kappa score: 0.027 ------ Step 3/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 460, 31 GAN fn, tp: 7, 6 GAN f1 score: 0.240 GAN cohens kappa score: 0.210 -> test with 'LR' LR tn, fp: 436, 55 LR fn, tp: 3, 10 LR f1 score: 0.256 LR cohens kappa score: 0.223 LR average precision score: 0.335 -> test with 'GB' GB tn, fp: 475, 16 GB fn, tp: 7, 6 GB f1 score: 0.343 GB cohens kappa score: 0.321 -> test with 'KNN' KNN tn, fp: 382, 109 KNN fn, tp: 8, 5 KNN f1 score: 0.079 KNN cohens kappa score: 0.034 ====== Step 4/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 4/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 357, 136 GAN fn, tp: 1, 14 GAN f1 score: 0.170 GAN cohens kappa score: 0.123 -> test with 'LR' LR tn, fp: 427, 66 LR fn, tp: 3, 12 LR f1 score: 0.258 LR cohens kappa score: 0.219 LR average precision score: 0.293 -> test with 'GB' GB tn, fp: 476, 17 GB fn, tp: 8, 7 GB f1 score: 0.359 GB cohens kappa score: 0.335 -> test with 'KNN' KNN tn, fp: 393, 100 KNN fn, tp: 11, 4 KNN f1 score: 0.067 KNN cohens kappa score: 0.016 ------ Step 4/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 464, 29 GAN fn, tp: 8, 7 GAN f1 score: 0.275 GAN cohens kappa score: 0.243 -> test with 'LR' LR tn, fp: 442, 51 LR fn, tp: 3, 12 LR f1 score: 0.308 LR cohens kappa score: 0.273 LR average precision score: 0.226 -> test with 'GB' GB tn, fp: 479, 14 GB fn, tp: 10, 5 GB f1 score: 0.294 GB cohens kappa score: 0.270 -> test with 'KNN' KNN tn, fp: 421, 72 KNN fn, tp: 7, 8 KNN f1 score: 0.168 KNN cohens kappa score: 0.125 ------ Step 4/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 413, 80 GAN fn, tp: 4, 11 GAN f1 score: 0.208 GAN cohens kappa score: 0.165 -> test with 'LR' LR tn, fp: 446, 47 LR fn, tp: 3, 12 LR f1 score: 0.324 LR cohens kappa score: 0.291 LR average precision score: 0.197 -> test with 'GB' GB tn, fp: 483, 10 GB fn, tp: 8, 7 GB f1 score: 0.437 GB cohens kappa score: 0.419 -> test with 'KNN' KNN tn, fp: 409, 84 KNN fn, tp: 9, 6 KNN f1 score: 0.114 KNN cohens kappa score: 0.067 ------ Step 4/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 313, 180 GAN fn, tp: 3, 12 GAN f1 score: 0.116 GAN cohens kappa score: 0.065 -> test with 'LR' LR tn, fp: 432, 61 LR fn, tp: 3, 12 LR f1 score: 0.273 LR cohens kappa score: 0.235 LR average precision score: 0.288 -> test with 'GB' GB tn, fp: 486, 7 GB fn, tp: 8, 7 GB f1 score: 0.483 GB cohens kappa score: 0.468 -> test with 'KNN' KNN tn, fp: 392, 101 KNN fn, tp: 9, 6 KNN f1 score: 0.098 KNN cohens kappa score: 0.049 ------ Step 4/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 361, 130 GAN fn, tp: 2, 11 GAN f1 score: 0.143 GAN cohens kappa score: 0.100 -> test with 'LR' LR tn, fp: 430, 61 LR fn, tp: 4, 9 LR f1 score: 0.217 LR cohens kappa score: 0.181 LR average precision score: 0.177 -> test with 'GB' GB tn, fp: 477, 14 GB fn, tp: 7, 6 GB f1 score: 0.364 GB cohens kappa score: 0.343 -> test with 'KNN' KNN tn, fp: 388, 103 KNN fn, tp: 6, 7 KNN f1 score: 0.114 KNN cohens kappa score: 0.071 ====== Step 5/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 5/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 451, 42 GAN fn, tp: 6, 9 GAN f1 score: 0.273 GAN cohens kappa score: 0.238 -> test with 'LR' LR tn, fp: 439, 54 LR fn, tp: 2, 13 LR f1 score: 0.317 LR cohens kappa score: 0.282 LR average precision score: 0.293 -> test with 'GB' GB tn, fp: 483, 10 GB fn, tp: 6, 9 GB f1 score: 0.529 GB cohens kappa score: 0.513 -> test with 'KNN' KNN tn, fp: 404, 89 KNN fn, tp: 10, 5 KNN f1 score: 0.092 KNN cohens kappa score: 0.043 ------ Step 5/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 415, 78 GAN fn, tp: 3, 12 GAN f1 score: 0.229 GAN cohens kappa score: 0.187 -> test with 'LR' LR tn, fp: 430, 63 LR fn, tp: 3, 12 LR f1 score: 0.267 LR cohens kappa score: 0.229 LR average precision score: 0.181 -> test with 'GB' GB tn, fp: 477, 16 GB fn, tp: 9, 6 GB f1 score: 0.324 GB cohens kappa score: 0.300 -> test with 'KNN' KNN tn, fp: 401, 92 KNN fn, tp: 11, 4 KNN f1 score: 0.072 KNN cohens kappa score: 0.022 ------ Step 5/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 445, 48 GAN fn, tp: 7, 8 GAN f1 score: 0.225 GAN cohens kappa score: 0.188 -> test with 'LR' LR tn, fp: 460, 33 LR fn, tp: 5, 10 LR f1 score: 0.345 LR cohens kappa score: 0.315 LR average precision score: 0.196 -> test with 'GB' GB tn, fp: 486, 7 GB fn, tp: 11, 4 GB f1 score: 0.308 GB cohens kappa score: 0.290 -> test with 'KNN' KNN tn, fp: 405, 88 KNN fn, tp: 10, 5 KNN f1 score: 0.093 KNN cohens kappa score: 0.044 ------ Step 5/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 379, 114 GAN fn, tp: 2, 13 GAN f1 score: 0.183 GAN cohens kappa score: 0.138 -> test with 'LR' LR tn, fp: 426, 67 LR fn, tp: 2, 13 LR f1 score: 0.274 LR cohens kappa score: 0.236 LR average precision score: 0.237 -> test with 'GB' GB tn, fp: 477, 16 GB fn, tp: 6, 9 GB f1 score: 0.450 GB cohens kappa score: 0.429 -> test with 'KNN' KNN tn, fp: 398, 95 KNN fn, tp: 7, 8 KNN f1 score: 0.136 KNN cohens kappa score: 0.089 ------ Step 5/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1912 synthetic samples -> test with GAN.predict GAN tn, fp: 380, 111 GAN fn, tp: 4, 9 GAN f1 score: 0.135 GAN cohens kappa score: 0.093 -> test with 'LR' LR tn, fp: 428, 63 LR fn, tp: 3, 10 LR f1 score: 0.233 LR cohens kappa score: 0.197 LR average precision score: 0.235 -> test with 'GB' GB tn, fp: 481, 10 GB fn, tp: 9, 4 GB f1 score: 0.296 GB cohens kappa score: 0.277 -> test with 'KNN' KNN tn, fp: 402, 89 KNN fn, tp: 8, 5 KNN f1 score: 0.093 KNN cohens kappa score: 0.050 ### Exercise is done. -----[ LR ]----- maximum: LR tn, fp: 460, 67 LR fn, tp: 6, 14 LR f1 score: 0.345 LR cohens kappa score: 0.315 LR average precision score: 0.468 average: LR tn, fp: 437.16, 55.44 LR fn, tp: 3.64, 10.96 LR f1 score: 0.272 LR cohens kappa score: 0.236 LR average precision score: 0.235 minimum: LR tn, fp: 426, 33 LR fn, tp: 1, 9 LR f1 score: 0.217 LR cohens kappa score: 0.181 LR average precision score: 0.128 -----[ GB ]----- maximum: GB tn, fp: 487, 21 GB fn, tp: 12, 9 GB f1 score: 0.529 GB cohens kappa score: 0.513 average: GB tn, fp: 480.44, 12.16 GB fn, tp: 8.84, 5.76 GB f1 score: 0.352 GB cohens kappa score: 0.331 minimum: GB tn, fp: 472, 6 GB fn, tp: 6, 2 GB f1 score: 0.148 GB cohens kappa score: 0.125 -----[ KNN ]----- maximum: KNN tn, fp: 423, 109 KNN fn, tp: 13, 9 KNN f1 score: 0.168 KNN cohens kappa score: 0.125 average: KNN tn, fp: 401.32, 91.28 KNN fn, tp: 9.04, 5.56 KNN f1 score: 0.100 KNN cohens kappa score: 0.053 minimum: KNN tn, fp: 382, 70 KNN fn, tp: 6, 2 KNN f1 score: 0.033 KNN cohens kappa score: -0.019 -----[ GAN ]----- maximum: GAN tn, fp: 493, 215 GAN fn, tp: 15, 15 GAN f1 score: 0.275 GAN cohens kappa score: 0.244 average: GAN tn, fp: 392.52, 100.08 GAN fn, tp: 4.56, 10.04 GAN f1 score: 0.179 GAN cohens kappa score: 0.139 minimum: GAN tn, fp: 278, 0 GAN fn, tp: 0, 0 GAN f1 score: 0.000 GAN cohens kappa score: 0.000