/////////////////////////////////////////// // Running convGAN-majority-full on folding_abalone9-18 /////////////////////////////////////////// Load 'data_input/folding_abalone9-18' 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 518 synthetic samples -> test with GAN.predict GAN tn, fp: 137, 1 GAN fn, tp: 4, 5 GAN f1 score: 0.667 GAN cohens kappa score: 0.649 -> test with 'LR' LR tn, fp: 127, 11 LR fn, tp: 1, 8 LR f1 score: 0.571 LR cohens kappa score: 0.533 LR average precision score: 0.892 -> test with 'GB' GB tn, fp: 134, 4 GB fn, tp: 6, 3 GB f1 score: 0.375 GB cohens kappa score: 0.340 -> test with 'KNN' KNN tn, fp: 136, 2 KNN fn, tp: 8, 1 KNN f1 score: 0.167 KNN cohens kappa score: 0.140 ------ Step 1/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 518 synthetic samples -> test with GAN.predict GAN tn, fp: 130, 8 GAN fn, tp: 8, 1 GAN f1 score: 0.111 GAN cohens kappa score: 0.053 -> test with 'LR' LR tn, fp: 133, 5 LR fn, tp: 4, 5 LR f1 score: 0.526 LR cohens kappa score: 0.494 LR average precision score: 0.544 -> test with 'GB' GB tn, fp: 134, 4 GB fn, tp: 8, 1 GB f1 score: 0.143 GB cohens kappa score: 0.104 -> test with 'KNN' KNN tn, fp: 135, 3 KNN fn, tp: 9, 0 KNN f1 score: 0.000 KNN cohens kappa score: -0.032 ------ Step 1/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 518 synthetic samples -> test with GAN.predict GAN tn, fp: 136, 2 GAN fn, tp: 8, 1 GAN f1 score: 0.167 GAN cohens kappa score: 0.140 -> test with 'LR' LR tn, fp: 131, 7 LR fn, tp: 1, 8 LR f1 score: 0.667 LR cohens kappa score: 0.639 LR average precision score: 0.802 -> test with 'GB' GB tn, fp: 136, 2 GB fn, tp: 8, 1 GB f1 score: 0.167 GB cohens kappa score: 0.140 -> test with 'KNN' KNN tn, fp: 138, 0 KNN fn, tp: 8, 1 KNN f1 score: 0.200 KNN cohens kappa score: 0.190 ------ Step 1/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 518 synthetic samples -> test with GAN.predict GAN tn, fp: 135, 3 GAN fn, tp: 6, 3 GAN f1 score: 0.400 GAN cohens kappa score: 0.369 -> test with 'LR' LR tn, fp: 131, 7 LR fn, tp: 3, 6 LR f1 score: 0.545 LR cohens kappa score: 0.510 LR average precision score: 0.593 -> test with 'GB' GB tn, fp: 137, 1 GB fn, tp: 7, 2 GB f1 score: 0.333 GB cohens kappa score: 0.312 -> test with 'KNN' KNN tn, fp: 138, 0 KNN fn, tp: 7, 2 KNN f1 score: 0.364 KNN cohens kappa score: 0.349 ------ Step 1/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 516 synthetic samples -> test with GAN.predict GAN tn, fp: 136, 1 GAN fn, tp: 4, 2 GAN f1 score: 0.444 GAN cohens kappa score: 0.428 -> test with 'LR' LR tn, fp: 130, 7 LR fn, tp: 2, 4 LR f1 score: 0.471 LR cohens kappa score: 0.440 LR average precision score: 0.480 -> test with 'GB' GB tn, fp: 137, 0 GB fn, tp: 5, 1 GB f1 score: 0.286 GB cohens kappa score: 0.277 -> test with 'KNN' KNN tn, fp: 137, 0 KNN fn, tp: 4, 2 KNN f1 score: 0.500 KNN cohens kappa score: 0.489 ====== Step 2/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 2/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 518 synthetic samples -> test with GAN.predict GAN tn, fp: 132, 6 GAN fn, tp: 8, 1 GAN f1 score: 0.125 GAN cohens kappa score: 0.075 -> test with 'LR' LR tn, fp: 130, 8 LR fn, tp: 1, 8 LR f1 score: 0.640 LR cohens kappa score: 0.609 LR average precision score: 0.632 -> test with 'GB' GB tn, fp: 136, 2 GB fn, tp: 7, 2 GB f1 score: 0.308 GB cohens kappa score: 0.281 -> test with 'KNN' KNN tn, fp: 135, 3 KNN fn, tp: 8, 1 KNN f1 score: 0.154 KNN cohens kappa score: 0.121 ------ Step 2/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 518 synthetic samples -> test with GAN.predict GAN tn, fp: 135, 3 GAN fn, tp: 7, 2 GAN f1 score: 0.286 GAN cohens kappa score: 0.253 -> test with 'LR' LR tn, fp: 132, 6 LR fn, tp: 2, 7 LR f1 score: 0.636 LR cohens kappa score: 0.608 LR average precision score: 0.759 -> test with 'GB' GB tn, fp: 136, 2 GB fn, tp: 6, 3 GB f1 score: 0.429 GB cohens kappa score: 0.402 -> test with 'KNN' KNN tn, fp: 137, 1 KNN fn, tp: 9, 0 KNN f1 score: 0.000 KNN cohens kappa score: -0.012 ------ Step 2/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 518 synthetic samples -> test with GAN.predict GAN tn, fp: 138, 0 GAN fn, tp: 5, 4 GAN f1 score: 0.615 GAN cohens kappa score: 0.600 -> test with 'LR' LR tn, fp: 133, 5 LR fn, tp: 2, 7 LR f1 score: 0.667 LR cohens kappa score: 0.642 LR average precision score: 0.730 -> test with 'GB' GB tn, fp: 134, 4 GB fn, tp: 7, 2 GB f1 score: 0.267 GB cohens kappa score: 0.229 -> test with 'KNN' KNN tn, fp: 138, 0 KNN fn, tp: 9, 0 KNN f1 score: 0.000 KNN cohens kappa score: 0.000 ------ Step 2/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 518 synthetic samples -> test with GAN.predict GAN tn, fp: 136, 2 GAN fn, tp: 7, 2 GAN f1 score: 0.308 GAN cohens kappa score: 0.281 -> test with 'LR' LR tn, fp: 133, 5 LR fn, tp: 3, 6 LR f1 score: 0.600 LR cohens kappa score: 0.571 LR average precision score: 0.739 -> test with 'GB' GB tn, fp: 134, 4 GB fn, tp: 6, 3 GB f1 score: 0.375 GB cohens kappa score: 0.340 -> test with 'KNN' KNN tn, fp: 137, 1 KNN fn, tp: 9, 0 KNN f1 score: 0.000 KNN cohens kappa score: -0.012 ------ Step 2/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 516 synthetic samples -> test with GAN.predict GAN tn, fp: 134, 3 GAN fn, tp: 3, 3 GAN f1 score: 0.500 GAN cohens kappa score: 0.478 -> test with 'LR' LR tn, fp: 129, 8 LR fn, tp: 1, 5 LR f1 score: 0.526 LR cohens kappa score: 0.497 LR average precision score: 0.657 -> test with 'GB' GB tn, fp: 136, 1 GB fn, tp: 3, 3 GB f1 score: 0.600 GB cohens kappa score: 0.586 -> test with 'KNN' KNN tn, fp: 136, 1 KNN fn, tp: 4, 2 KNN f1 score: 0.444 KNN cohens kappa score: 0.428 ====== Step 3/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 3/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 518 synthetic samples -> test with GAN.predict GAN tn, fp: 134, 4 GAN fn, tp: 5, 4 GAN f1 score: 0.471 GAN cohens kappa score: 0.438 -> test with 'LR' LR tn, fp: 130, 8 LR fn, tp: 5, 4 LR f1 score: 0.381 LR cohens kappa score: 0.334 LR average precision score: 0.547 -> test with 'GB' GB tn, fp: 134, 4 GB fn, tp: 8, 1 GB f1 score: 0.143 GB cohens kappa score: 0.104 -> test with 'KNN' KNN tn, fp: 136, 2 KNN fn, tp: 8, 1 KNN f1 score: 0.167 KNN cohens kappa score: 0.140 ------ Step 3/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 518 synthetic samples -> test with GAN.predict GAN tn, fp: 134, 4 GAN fn, tp: 5, 4 GAN f1 score: 0.471 GAN cohens kappa score: 0.438 -> test with 'LR' LR tn, fp: 133, 5 LR fn, tp: 0, 9 LR f1 score: 0.783 LR cohens kappa score: 0.765 LR average precision score: 0.906 -> test with 'GB' GB tn, fp: 136, 2 GB fn, tp: 7, 2 GB f1 score: 0.308 GB cohens kappa score: 0.281 -> test with 'KNN' KNN tn, fp: 134, 4 KNN fn, tp: 9, 0 KNN f1 score: 0.000 KNN cohens kappa score: -0.039 ------ Step 3/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 518 synthetic samples -> test with GAN.predict GAN tn, fp: 136, 2 GAN fn, tp: 7, 2 GAN f1 score: 0.308 GAN cohens kappa score: 0.281 -> test with 'LR' LR tn, fp: 134, 4 LR fn, tp: 4, 5 LR f1 score: 0.556 LR cohens kappa score: 0.527 LR average precision score: 0.632 -> test with 'GB' GB tn, fp: 137, 1 GB fn, tp: 7, 2 GB f1 score: 0.333 GB cohens kappa score: 0.312 -> test with 'KNN' KNN tn, fp: 137, 1 KNN fn, tp: 9, 0 KNN f1 score: 0.000 KNN cohens kappa score: -0.012 ------ Step 3/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 518 synthetic samples -> test with GAN.predict GAN tn, fp: 135, 3 GAN fn, tp: 6, 3 GAN f1 score: 0.400 GAN cohens kappa score: 0.369 -> test with 'LR' LR tn, fp: 122, 16 LR fn, tp: 2, 7 LR f1 score: 0.438 LR cohens kappa score: 0.383 LR average precision score: 0.659 -> test with 'GB' GB tn, fp: 137, 1 GB fn, tp: 6, 3 GB f1 score: 0.462 GB cohens kappa score: 0.440 -> test with 'KNN' KNN tn, fp: 136, 2 KNN fn, tp: 6, 3 KNN f1 score: 0.429 KNN cohens kappa score: 0.402 ------ Step 3/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 516 synthetic samples -> test with GAN.predict GAN tn, fp: 133, 4 GAN fn, tp: 5, 1 GAN f1 score: 0.182 GAN cohens kappa score: 0.149 -> test with 'LR' LR tn, fp: 129, 8 LR fn, tp: 2, 4 LR f1 score: 0.444 LR cohens kappa score: 0.412 LR average precision score: 0.528 -> test with 'GB' GB tn, fp: 133, 4 GB fn, tp: 5, 1 GB f1 score: 0.182 GB cohens kappa score: 0.149 -> test with 'KNN' KNN tn, fp: 136, 1 KNN fn, tp: 6, 0 KNN f1 score: 0.000 KNN cohens kappa score: -0.012 ====== Step 4/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 4/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 518 synthetic samples -> test with GAN.predict GAN tn, fp: 136, 2 GAN fn, tp: 6, 3 GAN f1 score: 0.429 GAN cohens kappa score: 0.402 -> test with 'LR' LR tn, fp: 130, 8 LR fn, tp: 5, 4 LR f1 score: 0.381 LR cohens kappa score: 0.334 LR average precision score: 0.532 -> test with 'GB' GB tn, fp: 136, 2 GB fn, tp: 6, 3 GB f1 score: 0.429 GB cohens kappa score: 0.402 -> test with 'KNN' KNN tn, fp: 138, 0 KNN fn, tp: 9, 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 -> create 518 synthetic samples -> test with GAN.predict GAN tn, fp: 136, 2 GAN fn, tp: 5, 4 GAN f1 score: 0.533 GAN cohens kappa score: 0.509 -> test with 'LR' LR tn, fp: 130, 8 LR fn, tp: 3, 6 LR f1 score: 0.522 LR cohens kappa score: 0.483 LR average precision score: 0.681 -> test with 'GB' GB tn, fp: 132, 6 GB fn, tp: 5, 4 GB f1 score: 0.421 GB cohens kappa score: 0.381 -> test with 'KNN' KNN tn, fp: 138, 0 KNN fn, tp: 8, 1 KNN f1 score: 0.200 KNN cohens kappa score: 0.190 ------ Step 4/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 518 synthetic samples -> test with GAN.predict GAN tn, fp: 137, 1 GAN fn, tp: 6, 3 GAN f1 score: 0.462 GAN cohens kappa score: 0.440 -> test with 'LR' LR tn, fp: 131, 7 LR fn, tp: 2, 7 LR f1 score: 0.609 LR cohens kappa score: 0.577 LR average precision score: 0.652 -> test with 'GB' GB tn, fp: 137, 1 GB fn, tp: 7, 2 GB f1 score: 0.333 GB cohens kappa score: 0.312 -> test with 'KNN' KNN tn, fp: 138, 0 KNN fn, tp: 9, 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 -> create 518 synthetic samples -> test with GAN.predict GAN tn, fp: 133, 5 GAN fn, tp: 6, 3 GAN f1 score: 0.353 GAN cohens kappa score: 0.313 -> test with 'LR' LR tn, fp: 130, 8 LR fn, tp: 0, 9 LR f1 score: 0.692 LR cohens kappa score: 0.666 LR average precision score: 0.906 -> test with 'GB' GB tn, fp: 136, 2 GB fn, tp: 8, 1 GB f1 score: 0.167 GB cohens kappa score: 0.140 -> test with 'KNN' KNN tn, fp: 135, 3 KNN fn, tp: 9, 0 KNN f1 score: 0.000 KNN cohens kappa score: -0.032 ------ Step 4/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 516 synthetic samples -> test with GAN.predict GAN tn, fp: 134, 3 GAN fn, tp: 3, 3 GAN f1 score: 0.500 GAN cohens kappa score: 0.478 -> test with 'LR' LR tn, fp: 132, 5 LR fn, tp: 2, 4 LR f1 score: 0.533 LR cohens kappa score: 0.509 LR average precision score: 0.623 -> test with 'GB' GB tn, fp: 136, 1 GB fn, tp: 5, 1 GB f1 score: 0.250 GB cohens kappa score: 0.234 -> test with 'KNN' KNN tn, fp: 136, 1 KNN fn, tp: 5, 1 KNN f1 score: 0.250 KNN cohens kappa score: 0.234 ====== Step 5/5 ======= -> Shuffling data -> Spliting data to slices ------ Step 5/5: Slice 1/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 518 synthetic samples -> test with GAN.predict GAN tn, fp: 133, 5 GAN fn, tp: 6, 3 GAN f1 score: 0.353 GAN cohens kappa score: 0.313 -> test with 'LR' LR tn, fp: 131, 7 LR fn, tp: 2, 7 LR f1 score: 0.609 LR cohens kappa score: 0.577 LR average precision score: 0.705 -> test with 'GB' GB tn, fp: 136, 2 GB fn, tp: 8, 1 GB f1 score: 0.167 GB cohens kappa score: 0.140 -> test with 'KNN' KNN tn, fp: 136, 2 KNN fn, tp: 8, 1 KNN f1 score: 0.167 KNN cohens kappa score: 0.140 ------ Step 5/5: Slice 2/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 518 synthetic samples -> test with GAN.predict GAN tn, fp: 138, 0 GAN fn, tp: 7, 2 GAN f1 score: 0.364 GAN cohens kappa score: 0.349 -> test with 'LR' LR tn, fp: 130, 8 LR fn, tp: 2, 7 LR f1 score: 0.583 LR cohens kappa score: 0.549 LR average precision score: 0.717 -> test with 'GB' GB tn, fp: 136, 2 GB fn, tp: 5, 4 GB f1 score: 0.533 GB cohens kappa score: 0.509 -> test with 'KNN' KNN tn, fp: 138, 0 KNN fn, tp: 9, 0 KNN f1 score: 0.000 KNN cohens kappa score: 0.000 ------ Step 5/5: Slice 3/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 518 synthetic samples -> test with GAN.predict GAN tn, fp: 135, 3 GAN fn, tp: 7, 2 GAN f1 score: 0.286 GAN cohens kappa score: 0.253 -> test with 'LR' LR tn, fp: 130, 8 LR fn, tp: 4, 5 LR f1 score: 0.455 LR cohens kappa score: 0.412 LR average precision score: 0.537 -> test with 'GB' GB tn, fp: 136, 2 GB fn, tp: 8, 1 GB f1 score: 0.167 GB cohens kappa score: 0.140 -> test with 'KNN' KNN tn, fp: 136, 2 KNN fn, tp: 8, 1 KNN f1 score: 0.167 KNN cohens kappa score: 0.140 ------ Step 5/5: Slice 4/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 518 synthetic samples -> test with GAN.predict GAN tn, fp: 134, 4 GAN fn, tp: 6, 3 GAN f1 score: 0.375 GAN cohens kappa score: 0.340 -> test with 'LR' LR tn, fp: 131, 7 LR fn, tp: 1, 8 LR f1 score: 0.667 LR cohens kappa score: 0.639 LR average precision score: 0.891 -> test with 'GB' GB tn, fp: 138, 0 GB fn, tp: 6, 3 GB f1 score: 0.500 GB cohens kappa score: 0.484 -> test with 'KNN' KNN tn, fp: 136, 2 KNN fn, tp: 7, 2 KNN f1 score: 0.308 KNN cohens kappa score: 0.281 ------ Step 5/5: Slice 5/5 ------- -> Reset the GAN -> Train generator for synthetic samples -> create 516 synthetic samples -> test with GAN.predict GAN tn, fp: 136, 1 GAN fn, tp: 3, 3 GAN f1 score: 0.600 GAN cohens kappa score: 0.586 -> test with 'LR' LR tn, fp: 131, 6 LR fn, tp: 1, 5 LR f1 score: 0.588 LR cohens kappa score: 0.565 LR average precision score: 0.827 -> test with 'GB' GB tn, fp: 137, 0 GB fn, tp: 4, 2 GB f1 score: 0.500 GB cohens kappa score: 0.489 -> test with 'KNN' KNN tn, fp: 137, 0 KNN fn, tp: 4, 2 KNN f1 score: 0.500 KNN cohens kappa score: 0.489 ### Exercise is done. -----[ LR ]----- maximum: LR tn, fp: 134, 16 LR fn, tp: 5, 9 LR f1 score: 0.783 LR cohens kappa score: 0.765 LR average precision score: 0.906 average: LR tn, fp: 130.52, 7.28 LR fn, tp: 2.2, 6.2 LR f1 score: 0.564 LR cohens kappa score: 0.531 LR average precision score: 0.687 minimum: LR tn, fp: 122, 4 LR fn, tp: 0, 4 LR f1 score: 0.381 LR cohens kappa score: 0.334 LR average precision score: 0.480 -----[ GB ]----- maximum: GB tn, fp: 138, 6 GB fn, tp: 8, 4 GB f1 score: 0.600 GB cohens kappa score: 0.586 average: GB tn, fp: 135.64, 2.16 GB fn, tp: 6.32, 2.08 GB f1 score: 0.327 GB cohens kappa score: 0.301 minimum: GB tn, fp: 132, 0 GB fn, tp: 3, 1 GB f1 score: 0.143 GB cohens kappa score: 0.104 -----[ KNN ]----- maximum: KNN tn, fp: 138, 4 KNN fn, tp: 9, 3 KNN f1 score: 0.500 KNN cohens kappa score: 0.489 average: KNN tn, fp: 136.56, 1.24 KNN fn, tp: 7.56, 0.84 KNN f1 score: 0.161 KNN cohens kappa score: 0.143 minimum: KNN tn, fp: 134, 0 KNN fn, tp: 4, 0 KNN f1 score: 0.000 KNN cohens kappa score: -0.039 -----[ GAN ]----- maximum: GAN tn, fp: 138, 8 GAN fn, tp: 8, 5 GAN f1 score: 0.667 GAN cohens kappa score: 0.649 average: GAN tn, fp: 134.92, 2.88 GAN fn, tp: 5.72, 2.68 GAN f1 score: 0.388 GAN cohens kappa score: 0.360 minimum: GAN tn, fp: 130, 0 GAN fn, tp: 3, 1 GAN f1 score: 0.111 GAN cohens kappa score: 0.053