/////////////////////////////////////////// // Running convGAN-majority-full on folding_car_good /////////////////////////////////////////// Load 'data_input/folding_car_good' from pickle file 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 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 328, 4 GAN fn, tp: 2, 12 GAN f1 score: 0.800 GAN cohens kappa score: 0.791 -> test with 'LR' LR tn, fp: 179, 153 LR fn, tp: 5, 9 LR f1 score: 0.102 LR cohens kappa score: 0.030 LR average precision score: 0.070 -> test with 'GB' GB tn, fp: 330, 2 GB fn, tp: 2, 12 GB f1 score: 0.857 GB cohens kappa score: 0.851 -> test with 'KNN' KNN tn, fp: 297, 35 KNN fn, tp: 0, 14 KNN f1 score: 0.444 KNN cohens kappa score: 0.407 ------ Step 1/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 327, 5 GAN fn, tp: 4, 10 GAN f1 score: 0.690 GAN cohens kappa score: 0.676 -> test with 'LR' LR tn, fp: 190, 142 LR fn, tp: 1, 13 LR f1 score: 0.154 LR cohens kappa score: 0.086 LR average precision score: 0.090 -> test with 'GB' GB tn, fp: 330, 2 GB fn, tp: 5, 9 GB f1 score: 0.720 GB cohens kappa score: 0.710 -> test with 'KNN' KNN tn, fp: 291, 41 KNN fn, tp: 0, 14 KNN f1 score: 0.406 KNN cohens kappa score: 0.365 ------ Step 1/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 325, 7 GAN fn, tp: 3, 11 GAN f1 score: 0.688 GAN cohens kappa score: 0.673 -> test with 'LR' LR tn, fp: 186, 146 LR fn, tp: 5, 9 LR f1 score: 0.107 LR cohens kappa score: 0.035 LR average precision score: 0.061 -> test with 'GB' GB tn, fp: 331, 1 GB fn, tp: 3, 11 GB f1 score: 0.846 GB cohens kappa score: 0.840 -> test with 'KNN' KNN tn, fp: 307, 25 KNN fn, tp: 0, 14 KNN f1 score: 0.528 KNN cohens kappa score: 0.498 ------ Step 1/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 324, 8 GAN fn, tp: 5, 9 GAN f1 score: 0.581 GAN cohens kappa score: 0.561 -> test with 'LR' LR tn, fp: 188, 144 LR fn, tp: 5, 9 LR f1 score: 0.108 LR cohens kappa score: 0.036 LR average precision score: 0.077 -> test with 'GB' GB tn, fp: 330, 2 GB fn, tp: 2, 12 GB f1 score: 0.857 GB cohens kappa score: 0.851 -> test with 'KNN' KNN tn, fp: 294, 38 KNN fn, tp: 0, 14 KNN f1 score: 0.424 KNN cohens kappa score: 0.385 ------ Step 1/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 326, 5 GAN fn, tp: 4, 9 GAN f1 score: 0.667 GAN cohens kappa score: 0.653 -> test with 'LR' LR tn, fp: 179, 152 LR fn, tp: 4, 9 LR f1 score: 0.103 LR cohens kappa score: 0.036 LR average precision score: 0.055 -> test with 'GB' GB tn, fp: 326, 5 GB fn, tp: 1, 12 GB f1 score: 0.800 GB cohens kappa score: 0.791 -> test with 'KNN' KNN tn, fp: 298, 33 KNN fn, tp: 1, 12 KNN f1 score: 0.414 KNN cohens kappa score: 0.377 ====== Step 2/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 2/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 325, 7 GAN fn, tp: 3, 11 GAN f1 score: 0.688 GAN cohens kappa score: 0.673 -> test with 'LR' LR tn, fp: 166, 166 LR fn, tp: 4, 10 LR f1 score: 0.105 LR cohens kappa score: 0.033 LR average precision score: 0.072 -> test with 'GB' GB tn, fp: 331, 1 GB fn, tp: 3, 11 GB f1 score: 0.846 GB cohens kappa score: 0.840 -> test with 'KNN' KNN tn, fp: 310, 22 KNN fn, tp: 1, 13 KNN f1 score: 0.531 KNN cohens kappa score: 0.502 ------ Step 2/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 319, 13 GAN fn, tp: 1, 13 GAN f1 score: 0.650 GAN cohens kappa score: 0.631 -> test with 'LR' LR tn, fp: 177, 155 LR fn, tp: 4, 10 LR f1 score: 0.112 LR cohens kappa score: 0.040 LR average precision score: 0.072 -> test with 'GB' GB tn, fp: 330, 2 GB fn, tp: 1, 13 GB f1 score: 0.897 GB cohens kappa score: 0.892 -> test with 'KNN' KNN tn, fp: 300, 32 KNN fn, tp: 0, 14 KNN f1 score: 0.467 KNN cohens kappa score: 0.431 ------ Step 2/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 322, 10 GAN fn, tp: 5, 9 GAN f1 score: 0.545 GAN cohens kappa score: 0.523 -> test with 'LR' LR tn, fp: 190, 142 LR fn, tp: 5, 9 LR f1 score: 0.109 LR cohens kappa score: 0.038 LR average precision score: 0.070 -> test with 'GB' GB tn, fp: 332, 0 GB fn, tp: 7, 7 GB f1 score: 0.667 GB cohens kappa score: 0.657 -> test with 'KNN' KNN tn, fp: 294, 38 KNN fn, tp: 1, 13 KNN f1 score: 0.400 KNN cohens kappa score: 0.359 ------ Step 2/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 323, 9 GAN fn, tp: 3, 11 GAN f1 score: 0.647 GAN cohens kappa score: 0.629 -> test with 'LR' LR tn, fp: 188, 144 LR fn, tp: 8, 6 LR f1 score: 0.073 LR cohens kappa score: -0.001 LR average precision score: 0.049 -> test with 'GB' GB tn, fp: 331, 1 GB fn, tp: 2, 12 GB f1 score: 0.889 GB cohens kappa score: 0.884 -> test with 'KNN' KNN tn, fp: 280, 52 KNN fn, tp: 0, 14 KNN f1 score: 0.350 KNN cohens kappa score: 0.303 ------ Step 2/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 326, 5 GAN fn, tp: 7, 6 GAN f1 score: 0.500 GAN cohens kappa score: 0.482 -> test with 'LR' LR tn, fp: 187, 144 LR fn, tp: 5, 8 LR f1 score: 0.097 LR cohens kappa score: 0.029 LR average precision score: 0.079 -> test with 'GB' GB tn, fp: 328, 3 GB fn, tp: 2, 11 GB f1 score: 0.815 GB cohens kappa score: 0.807 -> test with 'KNN' KNN tn, fp: 301, 30 KNN fn, tp: 0, 13 KNN f1 score: 0.464 KNN cohens kappa score: 0.431 ====== Step 3/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 3/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 326, 6 GAN fn, tp: 5, 9 GAN f1 score: 0.621 GAN cohens kappa score: 0.604 -> test with 'LR' LR tn, fp: 174, 158 LR fn, tp: 2, 12 LR f1 score: 0.130 LR cohens kappa score: 0.060 LR average precision score: 0.079 -> test with 'GB' GB tn, fp: 331, 1 GB fn, tp: 3, 11 GB f1 score: 0.846 GB cohens kappa score: 0.840 -> test with 'KNN' KNN tn, fp: 298, 34 KNN fn, tp: 0, 14 KNN f1 score: 0.452 KNN cohens kappa score: 0.415 ------ Step 3/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 323, 9 GAN fn, tp: 2, 12 GAN f1 score: 0.686 GAN cohens kappa score: 0.670 -> test with 'LR' LR tn, fp: 197, 135 LR fn, tp: 4, 10 LR f1 score: 0.126 LR cohens kappa score: 0.056 LR average precision score: 0.070 -> test with 'GB' GB tn, fp: 328, 4 GB fn, tp: 1, 13 GB f1 score: 0.839 GB cohens kappa score: 0.831 -> test with 'KNN' KNN tn, fp: 297, 35 KNN fn, tp: 0, 14 KNN f1 score: 0.444 KNN cohens kappa score: 0.407 ------ Step 3/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 328, 4 GAN fn, tp: 8, 6 GAN f1 score: 0.500 GAN cohens kappa score: 0.483 -> test with 'LR' LR tn, fp: 192, 140 LR fn, tp: 6, 8 LR f1 score: 0.099 LR cohens kappa score: 0.027 LR average precision score: 0.058 -> test with 'GB' GB tn, fp: 330, 2 GB fn, tp: 3, 11 GB f1 score: 0.815 GB cohens kappa score: 0.807 -> test with 'KNN' KNN tn, fp: 313, 19 KNN fn, tp: 2, 12 KNN f1 score: 0.533 KNN cohens kappa score: 0.506 ------ Step 3/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 326, 6 GAN fn, tp: 5, 9 GAN f1 score: 0.621 GAN cohens kappa score: 0.604 -> test with 'LR' LR tn, fp: 179, 153 LR fn, tp: 2, 12 LR f1 score: 0.134 LR cohens kappa score: 0.064 LR average precision score: 0.085 -> test with 'GB' GB tn, fp: 332, 0 GB fn, tp: 2, 12 GB f1 score: 0.923 GB cohens kappa score: 0.920 -> test with 'KNN' KNN tn, fp: 304, 28 KNN fn, tp: 0, 14 KNN f1 score: 0.500 KNN cohens kappa score: 0.468 ------ Step 3/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 317, 14 GAN fn, tp: 3, 10 GAN f1 score: 0.541 GAN cohens kappa score: 0.517 -> test with 'LR' LR tn, fp: 177, 154 LR fn, tp: 5, 8 LR f1 score: 0.091 LR cohens kappa score: 0.023 LR average precision score: 0.057 -> test with 'GB' GB tn, fp: 328, 3 GB fn, tp: 1, 12 GB f1 score: 0.857 GB cohens kappa score: 0.851 -> test with 'KNN' KNN tn, fp: 292, 39 KNN fn, tp: 0, 13 KNN f1 score: 0.400 KNN cohens kappa score: 0.361 ====== Step 4/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 4/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 324, 8 GAN fn, tp: 5, 9 GAN f1 score: 0.581 GAN cohens kappa score: 0.561 -> test with 'LR' LR tn, fp: 181, 151 LR fn, tp: 4, 10 LR f1 score: 0.114 LR cohens kappa score: 0.043 LR average precision score: 0.069 -> test with 'GB' GB tn, fp: 332, 0 GB fn, tp: 0, 14 GB f1 score: 1.000 GB cohens kappa score: 1.000 -> test with 'KNN' KNN tn, fp: 309, 23 KNN fn, tp: 0, 14 KNN f1 score: 0.549 KNN cohens kappa score: 0.521 ------ Step 4/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 323, 9 GAN fn, tp: 1, 13 GAN f1 score: 0.722 GAN cohens kappa score: 0.708 -> test with 'LR' LR tn, fp: 181, 151 LR fn, tp: 5, 9 LR f1 score: 0.103 LR cohens kappa score: 0.031 LR average precision score: 0.067 -> test with 'GB' GB tn, fp: 330, 2 GB fn, tp: 3, 11 GB f1 score: 0.815 GB cohens kappa score: 0.807 -> test with 'KNN' KNN tn, fp: 290, 42 KNN fn, tp: 0, 14 KNN f1 score: 0.400 KNN cohens kappa score: 0.358 ------ Step 4/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 321, 11 GAN fn, tp: 4, 10 GAN f1 score: 0.571 GAN cohens kappa score: 0.550 -> test with 'LR' LR tn, fp: 172, 160 LR fn, tp: 3, 11 LR f1 score: 0.119 LR cohens kappa score: 0.048 LR average precision score: 0.067 -> test with 'GB' GB tn, fp: 332, 0 GB fn, tp: 2, 12 GB f1 score: 0.923 GB cohens kappa score: 0.920 -> test with 'KNN' KNN tn, fp: 292, 40 KNN fn, tp: 0, 14 KNN f1 score: 0.412 KNN cohens kappa score: 0.371 ------ Step 4/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 324, 8 GAN fn, tp: 5, 9 GAN f1 score: 0.581 GAN cohens kappa score: 0.561 -> test with 'LR' LR tn, fp: 201, 131 LR fn, tp: 6, 8 LR f1 score: 0.105 LR cohens kappa score: 0.034 LR average precision score: 0.057 -> test with 'GB' GB tn, fp: 331, 1 GB fn, tp: 1, 13 GB f1 score: 0.929 GB cohens kappa score: 0.926 -> test with 'KNN' KNN tn, fp: 311, 21 KNN fn, tp: 1, 13 KNN f1 score: 0.542 KNN cohens kappa score: 0.514 ------ Step 4/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 326, 5 GAN fn, tp: 2, 11 GAN f1 score: 0.759 GAN cohens kappa score: 0.748 -> test with 'LR' LR tn, fp: 181, 150 LR fn, tp: 1, 12 LR f1 score: 0.137 LR cohens kappa score: 0.072 LR average precision score: 0.077 -> test with 'GB' GB tn, fp: 327, 4 GB fn, tp: 5, 8 GB f1 score: 0.640 GB cohens kappa score: 0.626 -> test with 'KNN' KNN tn, fp: 300, 31 KNN fn, tp: 1, 12 KNN f1 score: 0.429 KNN cohens kappa score: 0.393 ====== Step 5/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 5/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 324, 8 GAN fn, tp: 2, 12 GAN f1 score: 0.706 GAN cohens kappa score: 0.691 -> test with 'LR' LR tn, fp: 179, 153 LR fn, tp: 8, 6 LR f1 score: 0.069 LR cohens kappa score: -0.005 LR average precision score: 0.054 -> test with 'GB' GB tn, fp: 329, 3 GB fn, tp: 0, 14 GB f1 score: 0.903 GB cohens kappa score: 0.899 -> test with 'KNN' KNN tn, fp: 294, 38 KNN fn, tp: 0, 14 KNN f1 score: 0.424 KNN cohens kappa score: 0.385 ------ Step 5/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 328, 4 GAN fn, tp: 4, 10 GAN f1 score: 0.714 GAN cohens kappa score: 0.702 -> test with 'LR' LR tn, fp: 192, 140 LR fn, tp: 6, 8 LR f1 score: 0.099 LR cohens kappa score: 0.027 LR average precision score: 0.076 -> test with 'GB' GB tn, fp: 331, 1 GB fn, tp: 3, 11 GB f1 score: 0.846 GB cohens kappa score: 0.840 -> test with 'KNN' KNN tn, fp: 300, 32 KNN fn, tp: 0, 14 KNN f1 score: 0.467 KNN cohens kappa score: 0.431 ------ Step 5/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 323, 9 GAN fn, tp: 2, 12 GAN f1 score: 0.686 GAN cohens kappa score: 0.670 -> test with 'LR' LR tn, fp: 167, 165 LR fn, tp: 4, 10 LR f1 score: 0.106 LR cohens kappa score: 0.033 LR average precision score: 0.083 -> test with 'GB' GB tn, fp: 326, 6 GB fn, tp: 1, 13 GB f1 score: 0.788 GB cohens kappa score: 0.778 -> test with 'KNN' KNN tn, fp: 290, 42 KNN fn, tp: 0, 14 KNN f1 score: 0.400 KNN cohens kappa score: 0.358 ------ Step 5/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 325, 7 GAN fn, tp: 6, 8 GAN f1 score: 0.552 GAN cohens kappa score: 0.532 -> test with 'LR' LR tn, fp: 184, 148 LR fn, tp: 4, 10 LR f1 score: 0.116 LR cohens kappa score: 0.045 LR average precision score: 0.083 -> test with 'GB' GB tn, fp: 332, 0 GB fn, tp: 5, 9 GB f1 score: 0.783 GB cohens kappa score: 0.775 -> test with 'KNN' KNN tn, fp: 316, 16 KNN fn, tp: 1, 13 KNN f1 score: 0.605 KNN cohens kappa score: 0.582 ------ Step 5/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1272 synthetic samples -> test with GAN.predict GAN tn, fp: 324, 7 GAN fn, tp: 1, 12 GAN f1 score: 0.750 GAN cohens kappa score: 0.738 -> test with 'LR' LR tn, fp: 185, 146 LR fn, tp: 4, 9 LR f1 score: 0.107 LR cohens kappa score: 0.040 LR average precision score: 0.066 -> test with 'GB' GB tn, fp: 331, 0 GB fn, tp: 3, 10 GB f1 score: 0.870 GB cohens kappa score: 0.865 -> test with 'KNN' KNN tn, fp: 305, 26 KNN fn, tp: 0, 13 KNN f1 score: 0.500 KNN cohens kappa score: 0.470 ### Exercise is done. -----[ LR ]----- maximum: LR tn, fp: 201, 166 LR fn, tp: 8, 13 LR f1 score: 0.154 LR cohens kappa score: 0.086 LR average precision score: 0.090 average: LR tn, fp: 182.88, 148.92 LR fn, tp: 4.4, 9.4 LR f1 score: 0.109 LR cohens kappa score: 0.038 LR average precision score: 0.070 minimum: LR tn, fp: 166, 131 LR fn, tp: 1, 6 LR f1 score: 0.069 LR cohens kappa score: -0.005 LR average precision score: 0.049 -----[ GB ]----- maximum: GB tn, fp: 332, 6 GB fn, tp: 7, 14 GB f1 score: 1.000 GB cohens kappa score: 1.000 average: GB tn, fp: 329.96, 1.84 GB fn, tp: 2.44, 11.36 GB f1 score: 0.839 GB cohens kappa score: 0.832 minimum: GB tn, fp: 326, 0 GB fn, tp: 0, 7 GB f1 score: 0.640 GB cohens kappa score: 0.626 -----[ KNN ]----- maximum: KNN tn, fp: 316, 52 KNN fn, tp: 2, 14 KNN f1 score: 0.605 KNN cohens kappa score: 0.582 average: KNN tn, fp: 299.32, 32.48 KNN fn, tp: 0.32, 13.48 KNN f1 score: 0.459 KNN cohens kappa score: 0.424 minimum: KNN tn, fp: 280, 16 KNN fn, tp: 0, 12 KNN f1 score: 0.350 KNN cohens kappa score: 0.303 -----[ GAN ]----- maximum: GAN tn, fp: 328, 14 GAN fn, tp: 8, 13 GAN f1 score: 0.800 GAN cohens kappa score: 0.791 average: GAN tn, fp: 324.28, 7.52 GAN fn, tp: 3.68, 10.12 GAN f1 score: 0.642 GAN cohens kappa score: 0.625 minimum: GAN tn, fp: 317, 4 GAN fn, tp: 1, 6 GAN f1 score: 0.500 GAN cohens kappa score: 0.482