/////////////////////////////////////////// // Running convGAN-majority-full on folding_car-vgood /////////////////////////////////////////// Load 'data_input/folding_car-vgood' 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 1278 synthetic samples -> test with GAN.predict GAN tn, fp: 331, 2 GAN fn, tp: 7, 6 GAN f1 score: 0.571 GAN cohens kappa score: 0.559 -> test with 'LR' LR tn, fp: 303, 30 LR fn, tp: 1, 12 LR f1 score: 0.436 LR cohens kappa score: 0.402 LR average precision score: 0.363 -> test with 'GB' GB tn, fp: 333, 0 GB fn, tp: 0, 13 GB f1 score: 1.000 GB cohens kappa score: 1.000 -> test with 'KNN' KNN tn, fp: 333, 0 KNN fn, tp: 0, 13 KNN f1 score: 1.000 KNN cohens kappa score: 1.000 ------ Step 1/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1278 synthetic samples -> test with GAN.predict GAN tn, fp: 330, 3 GAN fn, tp: 5, 8 GAN f1 score: 0.667 GAN cohens kappa score: 0.655 -> test with 'LR' LR tn, fp: 297, 36 LR fn, tp: 1, 12 LR f1 score: 0.393 LR cohens kappa score: 0.355 LR average precision score: 0.296 -> test with 'GB' GB tn, fp: 333, 0 GB fn, tp: 1, 12 GB f1 score: 0.960 GB cohens kappa score: 0.959 -> test with 'KNN' KNN tn, fp: 323, 10 KNN fn, tp: 0, 13 KNN f1 score: 0.722 KNN cohens kappa score: 0.708 ------ Step 1/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1278 synthetic samples -> test with GAN.predict GAN tn, fp: 330, 3 GAN fn, tp: 8, 5 GAN f1 score: 0.476 GAN cohens kappa score: 0.461 -> test with 'LR' LR tn, fp: 294, 39 LR fn, tp: 0, 13 LR f1 score: 0.400 LR cohens kappa score: 0.362 LR average precision score: 0.392 -> test with 'GB' GB tn, fp: 333, 0 GB fn, tp: 1, 12 GB f1 score: 0.960 GB cohens kappa score: 0.959 -> test with 'KNN' KNN tn, fp: 325, 8 KNN fn, tp: 1, 12 KNN f1 score: 0.727 KNN cohens kappa score: 0.714 ------ Step 1/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1278 synthetic samples -> test with GAN.predict GAN tn, fp: 332, 1 GAN fn, tp: 5, 8 GAN f1 score: 0.727 GAN cohens kappa score: 0.719 -> test with 'LR' LR tn, fp: 295, 38 LR fn, tp: 1, 12 LR f1 score: 0.381 LR cohens kappa score: 0.342 LR average precision score: 0.374 -> test with 'GB' GB tn, fp: 333, 0 GB fn, tp: 0, 13 GB f1 score: 1.000 GB cohens kappa score: 1.000 -> test with 'KNN' KNN tn, fp: 329, 4 KNN fn, tp: 0, 13 KNN f1 score: 0.867 KNN cohens kappa score: 0.861 ------ Step 1/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1280 synthetic samples -> test with GAN.predict GAN tn, fp: 330, 1 GAN fn, tp: 7, 6 GAN f1 score: 0.600 GAN cohens kappa score: 0.589 -> test with 'LR' LR tn, fp: 306, 25 LR fn, tp: 3, 10 LR f1 score: 0.417 LR cohens kappa score: 0.383 LR average precision score: 0.433 -> test with 'GB' GB tn, fp: 329, 2 GB fn, tp: 0, 13 GB f1 score: 0.929 GB cohens kappa score: 0.926 -> test with 'KNN' KNN tn, fp: 329, 2 KNN fn, tp: 0, 13 KNN f1 score: 0.929 KNN cohens kappa score: 0.926 ====== Step 2/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 2/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1278 synthetic samples -> test with GAN.predict GAN tn, fp: 331, 2 GAN fn, tp: 6, 7 GAN f1 score: 0.636 GAN cohens kappa score: 0.625 -> test with 'LR' LR tn, fp: 305, 28 LR fn, tp: 5, 8 LR f1 score: 0.327 LR cohens kappa score: 0.287 LR average precision score: 0.284 -> test with 'GB' GB tn, fp: 333, 0 GB fn, tp: 0, 13 GB f1 score: 1.000 GB cohens kappa score: 1.000 -> test with 'KNN' KNN tn, fp: 325, 8 KNN fn, tp: 0, 13 KNN f1 score: 0.765 KNN cohens kappa score: 0.753 ------ Step 2/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1278 synthetic samples -> test with GAN.predict GAN tn, fp: 332, 1 GAN fn, tp: 8, 5 GAN f1 score: 0.526 GAN cohens kappa score: 0.515 -> test with 'LR' LR tn, fp: 294, 39 LR fn, tp: 0, 13 LR f1 score: 0.400 LR cohens kappa score: 0.362 LR average precision score: 0.329 -> test with 'GB' GB tn, fp: 333, 0 GB fn, tp: 0, 13 GB f1 score: 1.000 GB cohens kappa score: 1.000 -> test with 'KNN' KNN tn, fp: 326, 7 KNN fn, tp: 0, 13 KNN f1 score: 0.788 KNN cohens kappa score: 0.778 ------ Step 2/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1278 synthetic samples -> test with GAN.predict GAN tn, fp: 332, 1 GAN fn, tp: 9, 4 GAN f1 score: 0.444 GAN cohens kappa score: 0.433 -> test with 'LR' LR tn, fp: 303, 30 LR fn, tp: 2, 11 LR f1 score: 0.407 LR cohens kappa score: 0.372 LR average precision score: 0.334 -> test with 'GB' GB tn, fp: 333, 0 GB fn, tp: 0, 13 GB f1 score: 1.000 GB cohens kappa score: 1.000 -> test with 'KNN' KNN tn, fp: 331, 2 KNN fn, tp: 3, 10 KNN f1 score: 0.800 KNN cohens kappa score: 0.793 ------ Step 2/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1278 synthetic samples -> test with GAN.predict GAN tn, fp: 327, 6 GAN fn, tp: 5, 8 GAN f1 score: 0.593 GAN cohens kappa score: 0.576 -> test with 'LR' LR tn, fp: 301, 32 LR fn, tp: 0, 13 LR f1 score: 0.448 LR cohens kappa score: 0.414 LR average precision score: 0.295 -> test with 'GB' GB tn, fp: 333, 0 GB fn, tp: 3, 10 GB f1 score: 0.870 GB cohens kappa score: 0.865 -> test with 'KNN' KNN tn, fp: 326, 7 KNN fn, tp: 0, 13 KNN f1 score: 0.788 KNN cohens kappa score: 0.778 ------ Step 2/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1280 synthetic samples -> test with GAN.predict GAN tn, fp: 326, 5 GAN fn, tp: 5, 8 GAN f1 score: 0.615 GAN cohens kappa score: 0.600 -> test with 'LR' LR tn, fp: 298, 33 LR fn, tp: 1, 12 LR f1 score: 0.414 LR cohens kappa score: 0.377 LR average precision score: 0.549 -> test with 'GB' GB tn, fp: 331, 0 GB fn, tp: 0, 13 GB f1 score: 1.000 GB cohens kappa score: 1.000 -> test with 'KNN' KNN tn, fp: 331, 0 KNN fn, tp: 0, 13 KNN f1 score: 1.000 KNN cohens kappa score: 1.000 ====== Step 3/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 3/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1278 synthetic samples -> test with GAN.predict GAN tn, fp: 330, 3 GAN fn, tp: 5, 8 GAN f1 score: 0.667 GAN cohens kappa score: 0.655 -> test with 'LR' LR tn, fp: 301, 32 LR fn, tp: 2, 11 LR f1 score: 0.393 LR cohens kappa score: 0.356 LR average precision score: 0.309 -> test with 'GB' GB tn, fp: 333, 0 GB fn, tp: 2, 11 GB f1 score: 0.917 GB cohens kappa score: 0.914 -> test with 'KNN' KNN tn, fp: 329, 4 KNN fn, tp: 1, 12 KNN f1 score: 0.828 KNN cohens kappa score: 0.820 ------ Step 3/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1278 synthetic samples -> test with GAN.predict GAN tn, fp: 333, 0 GAN fn, tp: 5, 8 GAN f1 score: 0.762 GAN cohens kappa score: 0.755 -> test with 'LR' LR tn, fp: 302, 31 LR fn, tp: 0, 13 LR f1 score: 0.456 LR cohens kappa score: 0.423 LR average precision score: 0.431 -> test with 'GB' GB tn, fp: 332, 1 GB fn, tp: 0, 13 GB f1 score: 0.963 GB cohens kappa score: 0.961 -> test with 'KNN' KNN tn, fp: 329, 4 KNN fn, tp: 0, 13 KNN f1 score: 0.867 KNN cohens kappa score: 0.861 ------ Step 3/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1278 synthetic samples -> test with GAN.predict GAN tn, fp: 332, 1 GAN fn, tp: 5, 8 GAN f1 score: 0.727 GAN cohens kappa score: 0.719 -> test with 'LR' LR tn, fp: 289, 44 LR fn, tp: 0, 13 LR f1 score: 0.371 LR cohens kappa score: 0.330 LR average precision score: 0.322 -> test with 'GB' GB tn, fp: 333, 0 GB fn, tp: 0, 13 GB f1 score: 1.000 GB cohens kappa score: 1.000 -> test with 'KNN' KNN tn, fp: 325, 8 KNN fn, tp: 0, 13 KNN f1 score: 0.765 KNN cohens kappa score: 0.753 ------ Step 3/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1278 synthetic samples -> test with GAN.predict GAN tn, fp: 331, 2 GAN fn, tp: 3, 10 GAN f1 score: 0.800 GAN cohens kappa score: 0.793 -> test with 'LR' LR tn, fp: 297, 36 LR fn, tp: 0, 13 LR f1 score: 0.419 LR cohens kappa score: 0.383 LR average precision score: 0.382 -> test with 'GB' GB tn, fp: 333, 0 GB fn, tp: 0, 13 GB f1 score: 1.000 GB cohens kappa score: 1.000 -> test with 'KNN' KNN tn, fp: 328, 5 KNN fn, tp: 0, 13 KNN f1 score: 0.839 KNN cohens kappa score: 0.831 ------ Step 3/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1280 synthetic samples -> test with GAN.predict GAN tn, fp: 329, 2 GAN fn, tp: 5, 8 GAN f1 score: 0.696 GAN cohens kappa score: 0.685 -> test with 'LR' LR tn, fp: 298, 33 LR fn, tp: 3, 10 LR f1 score: 0.357 LR cohens kappa score: 0.318 LR average precision score: 0.380 -> test with 'GB' GB tn, fp: 331, 0 GB fn, tp: 0, 13 GB f1 score: 1.000 GB cohens kappa score: 1.000 -> test with 'KNN' KNN tn, fp: 328, 3 KNN fn, tp: 0, 13 KNN f1 score: 0.897 KNN cohens kappa score: 0.892 ====== Step 4/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 4/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1278 synthetic samples -> test with GAN.predict GAN tn, fp: 332, 1 GAN fn, tp: 5, 8 GAN f1 score: 0.727 GAN cohens kappa score: 0.719 -> test with 'LR' LR tn, fp: 302, 31 LR fn, tp: 1, 12 LR f1 score: 0.429 LR cohens kappa score: 0.394 LR average precision score: 0.419 -> test with 'GB' GB tn, fp: 333, 0 GB fn, tp: 0, 13 GB f1 score: 1.000 GB cohens kappa score: 1.000 -> test with 'KNN' KNN tn, fp: 327, 6 KNN fn, tp: 1, 12 KNN f1 score: 0.774 KNN cohens kappa score: 0.764 ------ Step 4/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1278 synthetic samples -> test with GAN.predict GAN tn, fp: 332, 1 GAN fn, tp: 5, 8 GAN f1 score: 0.727 GAN cohens kappa score: 0.719 -> test with 'LR' LR tn, fp: 294, 39 LR fn, tp: 1, 12 LR f1 score: 0.375 LR cohens kappa score: 0.335 LR average precision score: 0.510 -> test with 'GB' GB tn, fp: 333, 0 GB fn, tp: 0, 13 GB f1 score: 1.000 GB cohens kappa score: 1.000 -> test with 'KNN' KNN tn, fp: 327, 6 KNN fn, tp: 0, 13 KNN f1 score: 0.813 KNN cohens kappa score: 0.804 ------ Step 4/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1278 synthetic samples -> test with GAN.predict GAN tn, fp: 330, 3 GAN fn, tp: 2, 11 GAN f1 score: 0.815 GAN cohens kappa score: 0.807 -> test with 'LR' LR tn, fp: 295, 38 LR fn, tp: 0, 13 LR f1 score: 0.406 LR cohens kappa score: 0.368 LR average precision score: 0.321 -> test with 'GB' GB tn, fp: 332, 1 GB fn, tp: 0, 13 GB f1 score: 0.963 GB cohens kappa score: 0.961 -> test with 'KNN' KNN tn, fp: 326, 7 KNN fn, tp: 0, 13 KNN f1 score: 0.788 KNN cohens kappa score: 0.778 ------ Step 4/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1278 synthetic samples -> test with GAN.predict GAN tn, fp: 331, 2 GAN fn, tp: 9, 4 GAN f1 score: 0.421 GAN cohens kappa score: 0.407 -> test with 'LR' LR tn, fp: 302, 31 LR fn, tp: 3, 10 LR f1 score: 0.370 LR cohens kappa score: 0.332 LR average precision score: 0.276 -> test with 'GB' GB tn, fp: 333, 0 GB fn, tp: 0, 13 GB f1 score: 1.000 GB cohens kappa score: 1.000 -> test with 'KNN' KNN tn, fp: 327, 6 KNN fn, tp: 1, 12 KNN f1 score: 0.774 KNN cohens kappa score: 0.764 ------ Step 4/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1280 synthetic samples -> test with GAN.predict GAN tn, fp: 329, 2 GAN fn, tp: 5, 8 GAN f1 score: 0.696 GAN cohens kappa score: 0.685 -> test with 'LR' LR tn, fp: 301, 30 LR fn, tp: 1, 12 LR f1 score: 0.436 LR cohens kappa score: 0.402 LR average precision score: 0.327 -> test with 'GB' GB tn, fp: 331, 0 GB fn, tp: 0, 13 GB f1 score: 1.000 GB cohens kappa score: 1.000 -> test with 'KNN' KNN tn, fp: 324, 7 KNN fn, tp: 0, 13 KNN f1 score: 0.788 KNN cohens kappa score: 0.778 ====== Step 5/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 5/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1278 synthetic samples -> test with GAN.predict GAN tn, fp: 331, 2 GAN fn, tp: 4, 9 GAN f1 score: 0.750 GAN cohens kappa score: 0.741 -> test with 'LR' LR tn, fp: 290, 43 LR fn, tp: 0, 13 LR f1 score: 0.377 LR cohens kappa score: 0.336 LR average precision score: 0.292 -> test with 'GB' GB tn, fp: 333, 0 GB fn, tp: 0, 13 GB f1 score: 1.000 GB cohens kappa score: 1.000 -> test with 'KNN' KNN tn, fp: 329, 4 KNN fn, tp: 0, 13 KNN f1 score: 0.867 KNN cohens kappa score: 0.861 ------ Step 5/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1278 synthetic samples -> test with GAN.predict GAN tn, fp: 333, 0 GAN fn, tp: 8, 5 GAN f1 score: 0.556 GAN cohens kappa score: 0.546 -> test with 'LR' LR tn, fp: 312, 21 LR fn, tp: 3, 10 LR f1 score: 0.455 LR cohens kappa score: 0.424 LR average precision score: 0.338 -> test with 'GB' GB tn, fp: 333, 0 GB fn, tp: 0, 13 GB f1 score: 1.000 GB cohens kappa score: 1.000 -> test with 'KNN' KNN tn, fp: 330, 3 KNN fn, tp: 1, 12 KNN f1 score: 0.857 KNN cohens kappa score: 0.851 ------ Step 5/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1278 synthetic samples -> test with GAN.predict GAN tn, fp: 331, 2 GAN fn, tp: 8, 5 GAN f1 score: 0.500 GAN cohens kappa score: 0.486 -> test with 'LR' LR tn, fp: 312, 21 LR fn, tp: 3, 10 LR f1 score: 0.455 LR cohens kappa score: 0.424 LR average precision score: 0.338 -> test with 'GB' GB tn, fp: 333, 0 GB fn, tp: 0, 13 GB f1 score: 1.000 GB cohens kappa score: 1.000 -> test with 'KNN' KNN tn, fp: 333, 0 KNN fn, tp: 3, 10 KNN f1 score: 0.870 KNN cohens kappa score: 0.865 ------ Step 5/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1278 synthetic samples -> test with GAN.predict GAN tn, fp: 332, 1 GAN fn, tp: 8, 5 GAN f1 score: 0.526 GAN cohens kappa score: 0.515 -> test with 'LR' LR tn, fp: 298, 35 LR fn, tp: 1, 12 LR f1 score: 0.400 LR cohens kappa score: 0.362 LR average precision score: 0.294 -> test with 'GB' GB tn, fp: 333, 0 GB fn, tp: 0, 13 GB f1 score: 1.000 GB cohens kappa score: 1.000 -> test with 'KNN' KNN tn, fp: 329, 4 KNN fn, tp: 1, 12 KNN f1 score: 0.828 KNN cohens kappa score: 0.820 ------ Step 5/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 1280 synthetic samples -> test with GAN.predict GAN tn, fp: 330, 1 GAN fn, tp: 8, 5 GAN f1 score: 0.526 GAN cohens kappa score: 0.515 -> test with 'LR' LR tn, fp: 299, 32 LR fn, tp: 0, 13 LR f1 score: 0.448 LR cohens kappa score: 0.414 LR average precision score: 0.479 -> test with 'GB' GB tn, fp: 331, 0 GB fn, tp: 0, 13 GB f1 score: 1.000 GB cohens kappa score: 1.000 -> test with 'KNN' KNN tn, fp: 329, 2 KNN fn, tp: 2, 11 KNN f1 score: 0.846 KNN cohens kappa score: 0.840 ### Exercise is done. -----[ LR ]----- maximum: LR tn, fp: 312, 44 LR fn, tp: 5, 13 LR f1 score: 0.456 LR cohens kappa score: 0.424 LR average precision score: 0.549 average: LR tn, fp: 299.52, 33.08 LR fn, tp: 1.28, 11.72 LR f1 score: 0.407 LR cohens kappa score: 0.370 LR average precision score: 0.363 minimum: LR tn, fp: 289, 21 LR fn, tp: 0, 8 LR f1 score: 0.327 LR cohens kappa score: 0.287 LR average precision score: 0.276 -----[ GB ]----- maximum: GB tn, fp: 333, 2 GB fn, tp: 3, 13 GB f1 score: 1.000 GB cohens kappa score: 1.000 average: GB tn, fp: 332.44, 0.16 GB fn, tp: 0.28, 12.72 GB f1 score: 0.982 GB cohens kappa score: 0.982 minimum: GB tn, fp: 329, 0 GB fn, tp: 0, 10 GB f1 score: 0.870 GB cohens kappa score: 0.865 -----[ KNN ]----- maximum: KNN tn, fp: 333, 10 KNN fn, tp: 3, 13 KNN f1 score: 1.000 KNN cohens kappa score: 1.000 average: KNN tn, fp: 327.92, 4.68 KNN fn, tp: 0.56, 12.44 KNN f1 score: 0.831 KNN cohens kappa score: 0.824 minimum: KNN tn, fp: 323, 0 KNN fn, tp: 0, 10 KNN f1 score: 0.722 KNN cohens kappa score: 0.708 -----[ GAN ]----- maximum: GAN tn, fp: 333, 6 GAN fn, tp: 9, 11 GAN f1 score: 0.815 GAN cohens kappa score: 0.807 average: GAN tn, fp: 330.68, 1.92 GAN fn, tp: 6.0, 7.0 GAN f1 score: 0.630 GAN cohens kappa score: 0.619 minimum: GAN tn, fp: 326, 0 GAN fn, tp: 2, 4 GAN f1 score: 0.421 GAN cohens kappa score: 0.407