/////////////////////////////////////////// // Running SimpleGAN 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 Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227059 synthetic samples -> test with 'LR' LR tn, fp: 56821, 42 LR fn, tp: 35, 64 LR f1 score: 0.624 LR cohens kappa score: 0.624 LR average precision score: 0.489 -> test with 'GB' GB tn, fp: 56848, 15 GB fn, tp: 32, 67 GB f1 score: 0.740 GB cohens kappa score: 0.740 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 96, 3 KNN f1 score: 0.059 KNN cohens kappa score: 0.059 ------ Step 1/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227059 synthetic samples -> test with 'LR' LR tn, fp: 56853, 10 LR fn, tp: 30, 69 LR f1 score: 0.775 LR cohens kappa score: 0.775 LR average precision score: 0.707 -> test with 'GB' GB tn, fp: 56856, 7 GB fn, tp: 23, 76 GB f1 score: 0.835 GB cohens kappa score: 0.835 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 97, 2 KNN f1 score: 0.040 KNN cohens kappa score: 0.040 ------ Step 1/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227059 synthetic samples -> test with 'LR' LR tn, fp: 56852, 11 LR fn, tp: 36, 63 LR f1 score: 0.728 LR cohens kappa score: 0.728 LR average precision score: 0.640 -> test with 'GB' GB tn, fp: 56848, 15 GB fn, tp: 24, 75 GB f1 score: 0.794 GB cohens kappa score: 0.793 -> test with 'KNN' KNN tn, fp: 56862, 1 KNN fn, tp: 98, 1 KNN f1 score: 0.020 KNN cohens kappa score: 0.020 ------ Step 1/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227059 synthetic samples -> test with 'LR' LR tn, fp: 56847, 16 LR fn, tp: 30, 69 LR f1 score: 0.750 LR cohens kappa score: 0.750 LR average precision score: 0.743 -> test with 'GB' GB tn, fp: 56851, 12 GB fn, tp: 24, 75 GB f1 score: 0.806 GB cohens kappa score: 0.806 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 98, 1 KNN f1 score: 0.020 KNN cohens kappa score: 0.020 ------ Step 1/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227056 synthetic samples -> test with 'LR' LR tn, fp: 56854, 9 LR fn, tp: 37, 59 LR f1 score: 0.720 LR cohens kappa score: 0.719 LR average precision score: 0.738 -> test with 'GB' GB tn, fp: 56845, 18 GB fn, tp: 25, 71 GB f1 score: 0.768 GB cohens kappa score: 0.767 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 96, 0 KNN f1 score: 0.000 KNN cohens kappa score: 0.000 ====== Step 2/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 2/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227059 synthetic samples -> test with 'LR' LR tn, fp: 56835, 28 LR fn, tp: 40, 59 LR f1 score: 0.634 LR cohens kappa score: 0.634 LR average precision score: 0.612 -> test with 'GB' GB tn, fp: 56848, 15 GB fn, tp: 20, 79 GB f1 score: 0.819 GB cohens kappa score: 0.818 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 97, 2 KNN f1 score: 0.040 KNN cohens kappa score: 0.040 ------ Step 2/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227059 synthetic samples -> test with 'LR' LR tn, fp: 56842, 21 LR fn, tp: 34, 65 LR f1 score: 0.703 LR cohens kappa score: 0.702 LR average precision score: 0.606 -> test with 'GB' GB tn, fp: 56850, 13 GB fn, tp: 24, 75 GB f1 score: 0.802 GB cohens kappa score: 0.802 -> test with 'KNN' KNN tn, fp: 56862, 1 KNN fn, tp: 98, 1 KNN f1 score: 0.020 KNN cohens kappa score: 0.020 ------ Step 2/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227059 synthetic samples -> test with 'LR' LR tn, fp: 56847, 16 LR fn, tp: 40, 59 LR f1 score: 0.678 LR cohens kappa score: 0.678 LR average precision score: 0.610 -> test with 'GB' GB tn, fp: 56854, 9 GB fn, tp: 27, 72 GB f1 score: 0.800 GB cohens kappa score: 0.800 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 97, 2 KNN f1 score: 0.040 KNN cohens kappa score: 0.040 ------ Step 2/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227059 synthetic samples -> test with 'LR' LR tn, fp: 56850, 13 LR fn, tp: 43, 56 LR f1 score: 0.667 LR cohens kappa score: 0.666 LR average precision score: 0.630 -> test with 'GB' GB tn, fp: 56859, 4 GB fn, tp: 25, 74 GB f1 score: 0.836 GB cohens kappa score: 0.836 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 99, 0 KNN f1 score: 0.000 KNN cohens kappa score: 0.000 ------ Step 2/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227056 synthetic samples -> test with 'LR' LR tn, fp: 56849, 14 LR fn, tp: 31, 65 LR f1 score: 0.743 LR cohens kappa score: 0.742 LR average precision score: 0.683 -> test with 'GB' GB tn, fp: 56851, 12 GB fn, tp: 22, 74 GB f1 score: 0.813 GB cohens kappa score: 0.813 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 95, 1 KNN f1 score: 0.021 KNN cohens kappa score: 0.021 ====== Step 3/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 3/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227059 synthetic samples -> test with 'LR' LR tn, fp: 56844, 19 LR fn, tp: 42, 57 LR f1 score: 0.651 LR cohens kappa score: 0.651 LR average precision score: 0.595 -> test with 'GB' GB tn, fp: 56838, 25 GB fn, tp: 27, 72 GB f1 score: 0.735 GB cohens kappa score: 0.734 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 97, 2 KNN f1 score: 0.040 KNN cohens kappa score: 0.040 ------ Step 3/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227059 synthetic samples -> test with 'LR' LR tn, fp: 56846, 17 LR fn, tp: 30, 69 LR f1 score: 0.746 LR cohens kappa score: 0.746 LR average precision score: 0.671 -> test with 'GB' GB tn, fp: 56850, 13 GB fn, tp: 24, 75 GB f1 score: 0.802 GB cohens kappa score: 0.802 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 99, 0 KNN f1 score: 0.000 KNN cohens kappa score: 0.000 ------ Step 3/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227059 synthetic samples -> test with 'LR' LR tn, fp: 56850, 13 LR fn, tp: 33, 66 LR f1 score: 0.742 LR cohens kappa score: 0.741 LR average precision score: 0.680 -> test with 'GB' GB tn, fp: 56861, 2 GB fn, tp: 22, 77 GB f1 score: 0.865 GB cohens kappa score: 0.865 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 98, 1 KNN f1 score: 0.020 KNN cohens kappa score: 0.020 ------ Step 3/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227059 synthetic samples -> test with 'LR' LR tn, fp: 56852, 11 LR fn, tp: 43, 56 LR f1 score: 0.675 LR cohens kappa score: 0.674 LR average precision score: 0.674 -> test with 'GB' GB tn, fp: 56857, 6 GB fn, tp: 27, 72 GB f1 score: 0.814 GB cohens kappa score: 0.813 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 97, 2 KNN f1 score: 0.040 KNN cohens kappa score: 0.040 ------ Step 3/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227056 synthetic samples -> test with 'LR' LR tn, fp: 56858, 5 LR fn, tp: 35, 61 LR f1 score: 0.753 LR cohens kappa score: 0.753 LR average precision score: 0.702 -> test with 'GB' GB tn, fp: 56855, 8 GB fn, tp: 26, 70 GB f1 score: 0.805 GB cohens kappa score: 0.804 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 95, 1 KNN f1 score: 0.021 KNN cohens kappa score: 0.021 ====== Step 4/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 4/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227059 synthetic samples -> test with 'LR' LR tn, fp: 56847, 16 LR fn, tp: 33, 66 LR f1 score: 0.729 LR cohens kappa score: 0.729 LR average precision score: 0.676 -> test with 'GB' GB tn, fp: 56850, 13 GB fn, tp: 23, 76 GB f1 score: 0.809 GB cohens kappa score: 0.808 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 99, 0 KNN f1 score: 0.000 KNN cohens kappa score: 0.000 ------ Step 4/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227059 synthetic samples -> test with 'LR' LR tn, fp: 56843, 20 LR fn, tp: 35, 64 LR f1 score: 0.699 LR cohens kappa score: 0.699 LR average precision score: 0.585 -> test with 'GB' GB tn, fp: 56848, 15 GB fn, tp: 23, 76 GB f1 score: 0.800 GB cohens kappa score: 0.800 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 98, 1 KNN f1 score: 0.020 KNN cohens kappa score: 0.020 ------ Step 4/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227059 synthetic samples -> test with 'LR' LR tn, fp: 56853, 10 LR fn, tp: 32, 67 LR f1 score: 0.761 LR cohens kappa score: 0.761 LR average precision score: 0.700 -> test with 'GB' GB tn, fp: 56853, 10 GB fn, tp: 25, 74 GB f1 score: 0.809 GB cohens kappa score: 0.808 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 99, 0 KNN f1 score: 0.000 KNN cohens kappa score: 0.000 ------ Step 4/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227059 synthetic samples -> test with 'LR' LR tn, fp: 56856, 7 LR fn, tp: 32, 67 LR f1 score: 0.775 LR cohens kappa score: 0.774 LR average precision score: 0.721 -> test with 'GB' GB tn, fp: 56861, 2 GB fn, tp: 18, 81 GB f1 score: 0.890 GB cohens kappa score: 0.890 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 99, 0 KNN f1 score: 0.000 KNN cohens kappa score: 0.000 ------ Step 4/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227056 synthetic samples -> test with 'LR' LR tn, fp: 56855, 8 LR fn, tp: 45, 51 LR f1 score: 0.658 LR cohens kappa score: 0.658 LR average precision score: 0.653 -> test with 'GB' GB tn, fp: 56850, 13 GB fn, tp: 26, 70 GB f1 score: 0.782 GB cohens kappa score: 0.782 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 93, 3 KNN f1 score: 0.061 KNN cohens kappa score: 0.061 ====== Step 5/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 5/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227059 synthetic samples -> test with 'LR' LR tn, fp: 56849, 14 LR fn, tp: 38, 61 LR f1 score: 0.701 LR cohens kappa score: 0.701 LR average precision score: 0.660 -> test with 'GB' GB tn, fp: 56853, 10 GB fn, tp: 30, 69 GB f1 score: 0.775 GB cohens kappa score: 0.775 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 97, 2 KNN f1 score: 0.040 KNN cohens kappa score: 0.040 ------ Step 5/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227059 synthetic samples -> test with 'LR' LR tn, fp: 56856, 7 LR fn, tp: 36, 63 LR f1 score: 0.746 LR cohens kappa score: 0.745 LR average precision score: 0.778 -> test with 'GB' GB tn, fp: 56859, 4 GB fn, tp: 18, 81 GB f1 score: 0.880 GB cohens kappa score: 0.880 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 98, 1 KNN f1 score: 0.020 KNN cohens kappa score: 0.020 ------ Step 5/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227059 synthetic samples -> test with 'LR' LR tn, fp: 56853, 10 LR fn, tp: 33, 66 LR f1 score: 0.754 LR cohens kappa score: 0.754 LR average precision score: 0.652 -> test with 'GB' GB tn, fp: 56848, 15 GB fn, tp: 24, 75 GB f1 score: 0.794 GB cohens kappa score: 0.793 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 96, 3 KNN f1 score: 0.059 KNN cohens kappa score: 0.059 ------ Step 5/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227059 synthetic samples -> test with 'LR' LR tn, fp: 56856, 7 LR fn, tp: 40, 59 LR f1 score: 0.715 LR cohens kappa score: 0.715 LR average precision score: 0.729 -> test with 'GB' GB tn, fp: 56859, 4 GB fn, tp: 21, 78 GB f1 score: 0.862 GB cohens kappa score: 0.862 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 99, 0 KNN f1 score: 0.000 KNN cohens kappa score: 0.000 ------ Step 5/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples Epoch 1/3 Epoch 2/3 Epoch 3/3 -> create 227056 synthetic samples -> test with 'LR' LR tn, fp: 56850, 13 LR fn, tp: 37, 59 LR f1 score: 0.702 LR cohens kappa score: 0.702 LR average precision score: 0.628 -> test with 'GB' GB tn, fp: 56844, 19 GB fn, tp: 27, 69 GB f1 score: 0.750 GB cohens kappa score: 0.750 -> test with 'KNN' KNN tn, fp: 56863, 0 KNN fn, tp: 96, 0 KNN f1 score: 0.000 KNN cohens kappa score: 0.000 ### Exercise is done. -----[ LR ]----- maximum: LR tn, fp: 56858, 42 LR fn, tp: 45, 69 LR f1 score: 0.775 LR cohens kappa score: 0.775 LR average precision score: 0.778 average: LR tn, fp: 56848.72, 14.28 LR fn, tp: 36.0, 62.4 LR f1 score: 0.713 LR cohens kappa score: 0.713 LR average precision score: 0.662 minimum: LR tn, fp: 56821, 5 LR fn, tp: 30, 51 LR f1 score: 0.624 LR cohens kappa score: 0.624 LR average precision score: 0.489 -----[ GB ]----- maximum: GB tn, fp: 56861, 25 GB fn, tp: 32, 81 GB f1 score: 0.890 GB cohens kappa score: 0.890 average: GB tn, fp: 56851.84, 11.16 GB fn, tp: 24.28, 74.12 GB f1 score: 0.807 GB cohens kappa score: 0.807 minimum: GB tn, fp: 56838, 2 GB fn, tp: 18, 67 GB f1 score: 0.735 GB cohens kappa score: 0.734 -----[ KNN ]----- maximum: KNN tn, fp: 56863, 1 KNN fn, tp: 99, 3 KNN f1 score: 0.061 KNN cohens kappa score: 0.061 average: KNN tn, fp: 56862.92, 0.08 KNN fn, tp: 97.24, 1.16 KNN f1 score: 0.023 KNN cohens kappa score: 0.023 minimum: KNN tn, fp: 56862, 0 KNN fn, tp: 93, 0 KNN f1 score: 0.000 KNN cohens kappa score: 0.000