/////////////////////////////////////////// // Running convGAN-majority-full on kaggle_creditcard /////////////////////////////////////////// Load 'data_input/kaggle_creditcard' 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 227059 synthetic samples -> test with GAN.predict GAN tn, fp: 53161, 3702 GAN fn, tp: 18, 81 GAN f1 score: 0.042 GAN cohens kappa score: 0.038 -> test with 'LR' LR tn, fp: 53086, 3777 LR fn, tp: 20, 79 LR f1 score: 0.040 LR cohens kappa score: 0.037 LR average precision score: 0.518 -> test with 'GB' GB tn, fp: 56581, 282 GB fn, tp: 19, 80 GB f1 score: 0.347 GB cohens kappa score: 0.345 -> test with 'KNN' KNN tn, fp: 55211, 1652 KNN fn, tp: 71, 28 KNN f1 score: 0.031 KNN cohens kappa score: 0.028 ------ Step 1/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227059 synthetic samples -> test with GAN.predict GAN tn, fp: 43088, 13775 GAN fn, tp: 2, 97 GAN f1 score: 0.014 GAN cohens kappa score: 0.010 -> test with 'LR' LR tn, fp: 53787, 3076 LR fn, tp: 6, 93 LR f1 score: 0.057 LR cohens kappa score: 0.054 LR average precision score: 0.729 -> test with 'GB' GB tn, fp: 56521, 342 GB fn, tp: 10, 89 GB f1 score: 0.336 GB cohens kappa score: 0.334 -> test with 'KNN' KNN tn, fp: 55842, 1021 KNN fn, tp: 71, 28 KNN f1 score: 0.049 KNN cohens kappa score: 0.046 ------ Step 1/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227059 synthetic samples -> test with GAN.predict GAN tn, fp: 8231, 48632 GAN fn, tp: 2, 97 GAN f1 score: 0.004 GAN cohens kappa score: 0.001 -> test with 'LR' LR tn, fp: 54646, 2217 LR fn, tp: 6, 93 LR f1 score: 0.077 LR cohens kappa score: 0.074 LR average precision score: 0.707 -> test with 'GB' GB tn, fp: 56483, 380 GB fn, tp: 11, 88 GB f1 score: 0.310 GB cohens kappa score: 0.308 -> test with 'KNN' KNN tn, fp: 56049, 814 KNN fn, tp: 71, 28 KNN f1 score: 0.060 KNN cohens kappa score: 0.057 ------ Step 1/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227059 synthetic samples -> test with GAN.predict GAN tn, fp: 38275, 18588 GAN fn, tp: 0, 99 GAN f1 score: 0.011 GAN cohens kappa score: 0.007 -> test with 'LR' LR tn, fp: 55145, 1718 LR fn, tp: 6, 93 LR f1 score: 0.097 LR cohens kappa score: 0.094 LR average precision score: 0.764 -> test with 'GB' GB tn, fp: 56487, 376 GB fn, tp: 8, 91 GB f1 score: 0.322 GB cohens kappa score: 0.320 -> test with 'KNN' KNN tn, fp: 55182, 1681 KNN fn, tp: 59, 40 KNN f1 score: 0.044 KNN cohens kappa score: 0.041 ------ Step 1/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227056 synthetic samples -> test with GAN.predict GAN tn, fp: 56659, 204 GAN fn, tp: 18, 78 GAN f1 score: 0.413 GAN cohens kappa score: 0.411 -> test with 'LR' LR tn, fp: 54854, 2009 LR fn, tp: 7, 89 LR f1 score: 0.081 LR cohens kappa score: 0.078 LR average precision score: 0.838 -> test with 'GB' GB tn, fp: 56424, 439 GB fn, tp: 9, 87 GB f1 score: 0.280 GB cohens kappa score: 0.278 -> test with 'KNN' KNN tn, fp: 55912, 951 KNN fn, tp: 62, 34 KNN f1 score: 0.063 KNN cohens kappa score: 0.060 ====== Step 2/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 2/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227059 synthetic samples -> test with GAN.predict GAN tn, fp: 56413, 450 GAN fn, tp: 21, 78 GAN f1 score: 0.249 GAN cohens kappa score: 0.247 -> test with 'LR' LR tn, fp: 52754, 4109 LR fn, tp: 11, 88 LR f1 score: 0.041 LR cohens kappa score: 0.038 LR average precision score: 0.695 -> test with 'GB' GB tn, fp: 56439, 424 GB fn, tp: 10, 89 GB f1 score: 0.291 GB cohens kappa score: 0.289 -> test with 'KNN' KNN tn, fp: 55796, 1067 KNN fn, tp: 69, 30 KNN f1 score: 0.050 KNN cohens kappa score: 0.047 ------ Step 2/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227059 synthetic samples -> test with GAN.predict GAN tn, fp: 52914, 3949 GAN fn, tp: 8, 91 GAN f1 score: 0.044 GAN cohens kappa score: 0.041 -> test with 'LR' LR tn, fp: 54662, 2201 LR fn, tp: 8, 91 LR f1 score: 0.076 LR cohens kappa score: 0.073 LR average precision score: 0.667 -> test with 'GB' GB tn, fp: 56466, 397 GB fn, tp: 11, 88 GB f1 score: 0.301 GB cohens kappa score: 0.299 -> test with 'KNN' KNN tn, fp: 55755, 1108 KNN fn, tp: 66, 33 KNN f1 score: 0.053 KNN cohens kappa score: 0.050 ------ Step 2/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227059 synthetic samples -> test with GAN.predict GAN tn, fp: 56556, 307 GAN fn, tp: 24, 75 GAN f1 score: 0.312 GAN cohens kappa score: 0.310 -> test with 'LR' LR tn, fp: 54718, 2145 LR fn, tp: 8, 91 LR f1 score: 0.078 LR cohens kappa score: 0.075 LR average precision score: 0.715 -> test with 'GB' GB tn, fp: 56525, 338 GB fn, tp: 12, 87 GB f1 score: 0.332 GB cohens kappa score: 0.330 -> test with 'KNN' KNN tn, fp: 55500, 1363 KNN fn, tp: 60, 39 KNN f1 score: 0.052 KNN cohens kappa score: 0.049 ------ Step 2/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227059 synthetic samples -> test with GAN.predict GAN tn, fp: 3888, 52975 GAN fn, tp: 1, 98 GAN f1 score: 0.004 GAN cohens kappa score: 0.000 -> test with 'LR' LR tn, fp: 54534, 2329 LR fn, tp: 12, 87 LR f1 score: 0.069 LR cohens kappa score: 0.066 LR average precision score: 0.713 -> test with 'GB' GB tn, fp: 56507, 356 GB fn, tp: 11, 88 GB f1 score: 0.324 GB cohens kappa score: 0.322 -> test with 'KNN' KNN tn, fp: 56368, 495 KNN fn, tp: 72, 27 KNN f1 score: 0.087 KNN cohens kappa score: 0.084 ------ Step 2/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227056 synthetic samples -> test with GAN.predict GAN tn, fp: 54173, 2690 GAN fn, tp: 10, 86 GAN f1 score: 0.060 GAN cohens kappa score: 0.057 -> test with 'LR' LR tn, fp: 54782, 2081 LR fn, tp: 12, 84 LR f1 score: 0.074 LR cohens kappa score: 0.071 LR average precision score: 0.759 -> test with 'GB' GB tn, fp: 56481, 382 GB fn, tp: 15, 81 GB f1 score: 0.290 GB cohens kappa score: 0.288 -> test with 'KNN' KNN tn, fp: 55998, 865 KNN fn, tp: 74, 22 KNN f1 score: 0.045 KNN cohens kappa score: 0.042 ====== Step 3/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 3/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227059 synthetic samples -> test with GAN.predict GAN tn, fp: 56008, 855 GAN fn, tp: 16, 83 GAN f1 score: 0.160 GAN cohens kappa score: 0.157 -> test with 'LR' LR tn, fp: 53786, 3077 LR fn, tp: 8, 91 LR f1 score: 0.056 LR cohens kappa score: 0.053 LR average precision score: 0.659 -> test with 'GB' GB tn, fp: 56468, 395 GB fn, tp: 13, 86 GB f1 score: 0.297 GB cohens kappa score: 0.295 -> test with 'KNN' KNN tn, fp: 56174, 689 KNN fn, tp: 73, 26 KNN f1 score: 0.064 KNN cohens kappa score: 0.061 ------ Step 3/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227059 synthetic samples -> test with GAN.predict GAN tn, fp: 54949, 1914 GAN fn, tp: 10, 89 GAN f1 score: 0.085 GAN cohens kappa score: 0.082 -> test with 'LR' LR tn, fp: 54319, 2544 LR fn, tp: 7, 92 LR f1 score: 0.067 LR cohens kappa score: 0.064 LR average precision score: 0.632 -> test with 'GB' GB tn, fp: 56480, 383 GB fn, tp: 9, 90 GB f1 score: 0.315 GB cohens kappa score: 0.313 -> test with 'KNN' KNN tn, fp: 55640, 1223 KNN fn, tp: 63, 36 KNN f1 score: 0.053 KNN cohens kappa score: 0.050 ------ Step 3/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227059 synthetic samples -> test with GAN.predict GAN tn, fp: 56510, 353 GAN fn, tp: 19, 80 GAN f1 score: 0.301 GAN cohens kappa score: 0.299 -> test with 'LR' LR tn, fp: 54720, 2143 LR fn, tp: 10, 89 LR f1 score: 0.076 LR cohens kappa score: 0.073 LR average precision score: 0.705 -> test with 'GB' GB tn, fp: 56498, 365 GB fn, tp: 13, 86 GB f1 score: 0.313 GB cohens kappa score: 0.311 -> test with 'KNN' KNN tn, fp: 56393, 470 KNN fn, tp: 75, 24 KNN f1 score: 0.081 KNN cohens kappa score: 0.078 ------ Step 3/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227059 synthetic samples -> test with GAN.predict GAN tn, fp: 53475, 3388 GAN fn, tp: 6, 93 GAN f1 score: 0.052 GAN cohens kappa score: 0.049 -> test with 'LR' LR tn, fp: 54682, 2181 LR fn, tp: 8, 91 LR f1 score: 0.077 LR cohens kappa score: 0.074 LR average precision score: 0.783 -> test with 'GB' GB tn, fp: 56451, 412 GB fn, tp: 9, 90 GB f1 score: 0.300 GB cohens kappa score: 0.297 -> test with 'KNN' KNN tn, fp: 56008, 855 KNN fn, tp: 72, 27 KNN f1 score: 0.055 KNN cohens kappa score: 0.052 ------ Step 3/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227056 synthetic samples -> test with GAN.predict GAN tn, fp: 55179, 1684 GAN fn, tp: 10, 86 GAN f1 score: 0.092 GAN cohens kappa score: 0.089 -> test with 'LR' LR tn, fp: 55002, 1861 LR fn, tp: 5, 91 LR f1 score: 0.089 LR cohens kappa score: 0.086 LR average precision score: 0.720 -> test with 'GB' GB tn, fp: 56544, 319 GB fn, tp: 12, 84 GB f1 score: 0.337 GB cohens kappa score: 0.335 -> test with 'KNN' KNN tn, fp: 55746, 1117 KNN fn, tp: 63, 33 KNN f1 score: 0.053 KNN cohens kappa score: 0.050 ====== Step 4/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 4/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227059 synthetic samples -> test with GAN.predict GAN tn, fp: 56586, 277 GAN fn, tp: 30, 69 GAN f1 score: 0.310 GAN cohens kappa score: 0.308 -> test with 'LR' LR tn, fp: 54089, 2774 LR fn, tp: 3, 96 LR f1 score: 0.065 LR cohens kappa score: 0.062 LR average precision score: 0.678 -> test with 'GB' GB tn, fp: 56499, 364 GB fn, tp: 7, 92 GB f1 score: 0.332 GB cohens kappa score: 0.330 -> test with 'KNN' KNN tn, fp: 56138, 725 KNN fn, tp: 76, 23 KNN f1 score: 0.054 KNN cohens kappa score: 0.051 ------ Step 4/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227059 synthetic samples -> test with GAN.predict GAN tn, fp: 56474, 389 GAN fn, tp: 22, 77 GAN f1 score: 0.273 GAN cohens kappa score: 0.270 -> test with 'LR' LR tn, fp: 54363, 2500 LR fn, tp: 12, 87 LR f1 score: 0.065 LR cohens kappa score: 0.062 LR average precision score: 0.646 -> test with 'GB' GB tn, fp: 56536, 327 GB fn, tp: 12, 87 GB f1 score: 0.339 GB cohens kappa score: 0.337 -> test with 'KNN' KNN tn, fp: 55879, 984 KNN fn, tp: 78, 21 KNN f1 score: 0.038 KNN cohens kappa score: 0.035 ------ Step 4/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227059 synthetic samples -> test with GAN.predict GAN tn, fp: 54791, 2072 GAN fn, tp: 11, 88 GAN f1 score: 0.078 GAN cohens kappa score: 0.075 -> test with 'LR' LR tn, fp: 54784, 2079 LR fn, tp: 10, 89 LR f1 score: 0.079 LR cohens kappa score: 0.075 LR average precision score: 0.733 -> test with 'GB' GB tn, fp: 56502, 361 GB fn, tp: 12, 87 GB f1 score: 0.318 GB cohens kappa score: 0.316 -> test with 'KNN' KNN tn, fp: 55465, 1398 KNN fn, tp: 66, 33 KNN f1 score: 0.043 KNN cohens kappa score: 0.040 ------ Step 4/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227059 synthetic samples -> test with GAN.predict GAN tn, fp: 54843, 2020 GAN fn, tp: 8, 91 GAN f1 score: 0.082 GAN cohens kappa score: 0.079 -> test with 'LR' LR tn, fp: 54175, 2688 LR fn, tp: 9, 90 LR f1 score: 0.063 LR cohens kappa score: 0.059 LR average precision score: 0.741 -> test with 'GB' GB tn, fp: 56422, 441 GB fn, tp: 10, 89 GB f1 score: 0.283 GB cohens kappa score: 0.281 -> test with 'KNN' KNN tn, fp: 55767, 1096 KNN fn, tp: 62, 37 KNN f1 score: 0.060 KNN cohens kappa score: 0.057 ------ Step 4/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227056 synthetic samples -> test with GAN.predict GAN tn, fp: 32504, 24359 GAN fn, tp: 0, 96 GAN f1 score: 0.008 GAN cohens kappa score: 0.004 -> test with 'LR' LR tn, fp: 54932, 1931 LR fn, tp: 11, 85 LR f1 score: 0.080 LR cohens kappa score: 0.078 LR average precision score: 0.718 -> test with 'GB' GB tn, fp: 56523, 340 GB fn, tp: 15, 81 GB f1 score: 0.313 GB cohens kappa score: 0.311 -> test with 'KNN' KNN tn, fp: 56062, 801 KNN fn, tp: 71, 25 KNN f1 score: 0.054 KNN cohens kappa score: 0.051 ====== Step 5/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 5/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227059 synthetic samples -> test with GAN.predict GAN tn, fp: 56001, 862 GAN fn, tp: 20, 79 GAN f1 score: 0.152 GAN cohens kappa score: 0.149 -> test with 'LR' LR tn, fp: 53935, 2928 LR fn, tp: 17, 82 LR f1 score: 0.053 LR cohens kappa score: 0.050 LR average precision score: 0.616 -> test with 'GB' GB tn, fp: 56593, 270 GB fn, tp: 16, 83 GB f1 score: 0.367 GB cohens kappa score: 0.366 -> test with 'KNN' KNN tn, fp: 55419, 1444 KNN fn, tp: 66, 33 KNN f1 score: 0.042 KNN cohens kappa score: 0.039 ------ Step 5/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227059 synthetic samples -> test with GAN.predict GAN tn, fp: 56691, 172 GAN fn, tp: 30, 69 GAN f1 score: 0.406 GAN cohens kappa score: 0.404 -> test with 'LR' LR tn, fp: 54246, 2617 LR fn, tp: 5, 94 LR f1 score: 0.067 LR cohens kappa score: 0.064 LR average precision score: 0.746 -> test with 'GB' GB tn, fp: 56516, 347 GB fn, tp: 8, 91 GB f1 score: 0.339 GB cohens kappa score: 0.337 -> test with 'KNN' KNN tn, fp: 55656, 1207 KNN fn, tp: 68, 31 KNN f1 score: 0.046 KNN cohens kappa score: 0.043 ------ Step 5/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227059 synthetic samples -> test with GAN.predict GAN tn, fp: 53258, 3605 GAN fn, tp: 13, 86 GAN f1 score: 0.045 GAN cohens kappa score: 0.042 -> test with 'LR' LR tn, fp: 54527, 2336 LR fn, tp: 11, 88 LR f1 score: 0.070 LR cohens kappa score: 0.067 LR average precision score: 0.671 -> test with 'GB' GB tn, fp: 56481, 382 GB fn, tp: 13, 86 GB f1 score: 0.303 GB cohens kappa score: 0.301 -> test with 'KNN' KNN tn, fp: 56334, 529 KNN fn, tp: 74, 25 KNN f1 score: 0.077 KNN cohens kappa score: 0.074 ------ Step 5/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227059 synthetic samples -> test with GAN.predict GAN tn, fp: 55530, 1333 GAN fn, tp: 9, 90 GAN f1 score: 0.118 GAN cohens kappa score: 0.115 -> test with 'LR' LR tn, fp: 54654, 2209 LR fn, tp: 7, 92 LR f1 score: 0.077 LR cohens kappa score: 0.074 LR average precision score: 0.769 -> test with 'GB' GB tn, fp: 56451, 412 GB fn, tp: 10, 89 GB f1 score: 0.297 GB cohens kappa score: 0.295 -> test with 'KNN' KNN tn, fp: 55919, 944 KNN fn, tp: 67, 32 KNN f1 score: 0.060 KNN cohens kappa score: 0.057 ------ Step 5/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 227056 synthetic samples -> test with GAN.predict GAN tn, fp: 56221, 642 GAN fn, tp: 13, 83 GAN f1 score: 0.202 GAN cohens kappa score: 0.200 -> test with 'LR' LR tn, fp: 54469, 2394 LR fn, tp: 11, 85 LR f1 score: 0.066 LR cohens kappa score: 0.063 LR average precision score: 0.659 -> test with 'GB' GB tn, fp: 56476, 387 GB fn, tp: 9, 87 GB f1 score: 0.305 GB cohens kappa score: 0.303 -> test with 'KNN' KNN tn, fp: 55591, 1272 KNN fn, tp: 62, 34 KNN f1 score: 0.049 KNN cohens kappa score: 0.046 ### Exercise is done. -----[ LR ]----- maximum: LR tn, fp: 55145, 4109 LR fn, tp: 20, 96 LR f1 score: 0.097 LR cohens kappa score: 0.094 LR average precision score: 0.838 average: LR tn, fp: 54386.04, 2476.96 LR fn, tp: 9.2, 89.2 LR f1 score: 0.070 LR cohens kappa score: 0.066 LR average precision score: 0.703 minimum: LR tn, fp: 52754, 1718 LR fn, tp: 3, 79 LR f1 score: 0.040 LR cohens kappa score: 0.037 LR average precision score: 0.518 -----[ GB ]----- maximum: GB tn, fp: 56593, 441 GB fn, tp: 19, 92 GB f1 score: 0.367 GB cohens kappa score: 0.366 average: GB tn, fp: 56494.16, 368.84 GB fn, tp: 11.36, 87.04 GB f1 score: 0.316 GB cohens kappa score: 0.314 minimum: GB tn, fp: 56422, 270 GB fn, tp: 7, 80 GB f1 score: 0.280 GB cohens kappa score: 0.278 -----[ KNN ]----- maximum: KNN tn, fp: 56393, 1681 KNN fn, tp: 78, 40 KNN f1 score: 0.087 KNN cohens kappa score: 0.084 average: KNN tn, fp: 55832.16, 1030.84 KNN fn, tp: 68.44, 29.96 KNN f1 score: 0.054 KNN cohens kappa score: 0.052 minimum: KNN tn, fp: 55182, 470 KNN fn, tp: 59, 21 KNN f1 score: 0.031 KNN cohens kappa score: 0.028 -----[ GAN ]----- maximum: GAN tn, fp: 56691, 52975 GAN fn, tp: 30, 99 GAN f1 score: 0.413 GAN cohens kappa score: 0.411 average: GAN tn, fp: 49295.12, 7567.88 GAN fn, tp: 12.84, 85.56 GAN f1 score: 0.141 GAN cohens kappa score: 0.138 minimum: GAN tn, fp: 3888, 172 GAN fn, tp: 0, 69 GAN f1 score: 0.004 GAN cohens kappa score: 0.000