Results here
From IPRE Wiki
>>> train() Epoch # 10 | TSS Error: 8.3566 | Correct: 0.1458 | RMS Error: 0.4172 Epoch # 10, Layer = 'output' | Units: 0.1458 | Patterns: 0.0000 Epoch # 20 | TSS Error: 6.4829 | Correct: 0.4167 | RMS Error: 0.3675 Epoch # 20, Layer = 'output' | Units: 0.4167 | Patterns: 0.3333 Epoch # 30 | TSS Error: 4.8166 | Correct: 0.4375 | RMS Error: 0.3168 Epoch # 30, Layer = 'output' | Units: 0.4375 | Patterns: 0.1667 Epoch # 40 | TSS Error: 4.8478 | Correct: 0.4583 | RMS Error: 0.3178 Epoch # 40, Layer = 'output' | Units: 0.4583 | Patterns: 0.0833 Epoch # 50 | TSS Error: 5.7972 | Correct: 0.4583 | RMS Error: 0.3475 Epoch # 50, Layer = 'output' | Units: 0.4583 | Patterns: 0.0000 Epoch # 60 | TSS Error: 7.3393 | Correct: 0.5625 | RMS Error: 0.3910 Epoch # 60, Layer = 'output' | Units: 0.5625 | Patterns: 0.0000 Epoch # 70 | TSS Error: 4.6184 | Correct: 0.6042 | RMS Error: 0.3102 Epoch # 70, Layer = 'output' | Units: 0.6042 | Patterns: 0.0833 Epoch # 80 | TSS Error: 3.3136 | Correct: 0.5625 | RMS Error: 0.2627 Epoch # 80, Layer = 'output' | Units: 0.5625 | Patterns: 0.0000 Epoch # 90 | TSS Error: 6.5261 | Correct: 0.6250 | RMS Error: 0.3687 Epoch # 90, Layer = 'output' | Units: 0.6250 | Patterns: 0.2500 Epoch # 100 | TSS Error: 3.5050 | Correct: 0.6667 | RMS Error: 0.2702 Epoch # 100, Layer = 'output' | Units: 0.6667 | Patterns: 0.3333 Epoch # 110 | TSS Error: 4.6672 | Correct: 0.7083 | RMS Error: 0.3118 Epoch # 110, Layer = 'output' | Units: 0.7083 | Patterns: 0.3333 Epoch # 120 | TSS Error: 4.4930 | Correct: 0.7500 | RMS Error: 0.3059 Epoch # 120, Layer = 'output' | Units: 0.7500 | Patterns: 0.5000 Epoch # 130 | TSS Error: 0.4796 | Correct: 0.7917 | RMS Error: 0.1000 Epoch # 130, Layer = 'output' | Units: 0.7917 | Patterns: 0.5000 Epoch # 140 | TSS Error: 1.7687 | Correct: 0.7292 | RMS Error: 0.1920 Epoch # 140, Layer = 'output' | Units: 0.7292 | Patterns: 0.3333 Epoch # 150 | TSS Error: 4.0810 | Correct: 0.7083 | RMS Error: 0.2916 Epoch # 150, Layer = 'output' | Units: 0.7083 | Patterns: 0.3333 Epoch # 160 | TSS Error: 0.2887 | Correct: 0.7500 | RMS Error: 0.0776 Epoch # 160, Layer = 'output' | Units: 0.7500 | Patterns: 0.5000 Epoch # 170 | TSS Error: 0.3464 | Correct: 0.8125 | RMS Error: 0.0849 Epoch # 170, Layer = 'output' | Units: 0.8125 | Patterns: 0.5833 Epoch # 180 | TSS Error: 0.1056 | Correct: 0.9375 | RMS Error: 0.0469 Epoch # 180, Layer = 'output' | Units: 0.9375 | Patterns: 0.7500 ---------------------------------------------------- Final # 182 | TSS Error: 0.0644 | Correct: 1.0000 | RMS Error: 0.0366 Final # 182, Layer = 'output' | Units: 1.0000 | Patterns: 1.0000 ---------------------------------------------------- >>> test() Variations 1 2 3 4 5 6 7 8 j = 0 result = 1 j = 0 result = 1 j = 0 result = 1 j = 0 result = 1 j = 0 result = 1 j = 0 result = 1 j = 0 result = 1 j = 0 result = 1 j = 1 result = 2 j = 1 result = 2 j = 1 result = 2 j = 1 result = 2 j = 1 result = 2 j = 1 result = 2 j = 1 result = 2 j = 1 result = 2 j = 2 result = 3 j = 2 result = 3 j = 2 result = 3 j = 2 result = 3 j = 2 result = 3 j = 2 result = 3 j = 2 result = 3 j = 2 result = 3 j = 3 result = 4 j = 3 result = 4 j = 3 result = 4 j = 3 result = 4 j = 3 result = 4 j = 3 result = 4 j = 3 result = 4 j = 3 result = 4 >>> >>> train() Epoch # 10 | TSS Error: 8.7795 | Correct: 0.2500 | RMS Error: 0.4277 Epoch # 10, Layer = 'output' | Units: 0.2500 | Patterns: 0.0000 Epoch # 20 | TSS Error: 9.0301 | Correct: 0.5208 | RMS Error: 0.4337 Epoch # 20, Layer = 'output' | Units: 0.5208 | Patterns: 0.1667 Epoch # 30 | TSS Error: 8.8221 | Correct: 0.6042 | RMS Error: 0.4287 Epoch # 30, Layer = 'output' | Units: 0.6042 | Patterns: 0.2500 Epoch # 40 | TSS Error: 5.0652 | Correct: 0.5417 | RMS Error: 0.3248 Epoch # 40, Layer = 'output' | Units: 0.5417 | Patterns: 0.1667 Epoch # 50 | TSS Error: 8.1240 | Correct: 0.5625 | RMS Error: 0.4114 Epoch # 50, Layer = 'output' | Units: 0.5625 | Patterns: 0.2500 Epoch # 60 | TSS Error: 4.9242 | Correct: 0.5208 | RMS Error: 0.3203 Epoch # 60, Layer = 'output' | Units: 0.5208 | Patterns: 0.1667 Epoch # 70 | TSS Error: 3.8475 | Correct: 0.6458 | RMS Error: 0.2831 Epoch # 70, Layer = 'output' | Units: 0.6458 | Patterns: 0.3333 Epoch # 80 | TSS Error: 4.7748 | Correct: 0.6875 | RMS Error: 0.3154 Epoch # 80, Layer = 'output' | Units: 0.6875 | Patterns: 0.3333 Epoch # 90 | TSS Error: 4.1281 | Correct: 0.7083 | RMS Error: 0.2933 Epoch # 90, Layer = 'output' | Units: 0.7083 | Patterns: 0.4167 Epoch # 100 | TSS Error: 2.0875 | Correct: 0.8125 | RMS Error: 0.2085 Epoch # 100, Layer = 'output' | Units: 0.8125 | Patterns: 0.5000 Epoch # 110 | TSS Error: 1.8829 | Correct: 0.7500 | RMS Error: 0.1981 Epoch # 110, Layer = 'output' | Units: 0.7500 | Patterns: 0.5000 Epoch # 120 | TSS Error: 3.4107 | Correct: 0.8125 | RMS Error: 0.2666 Epoch # 120, Layer = 'output' | Units: 0.8125 | Patterns: 0.5000 Epoch # 130 | TSS Error: 0.1550 | Correct: 0.8750 | RMS Error: 0.0568 Epoch # 130, Layer = 'output' | Units: 0.8750 | Patterns: 0.7500 Epoch # 140 | TSS Error: 0.0925 | Correct: 1.0000 | RMS Error: 0.0439 Epoch # 140, Layer = 'output' | Units: 1.0000 | Patterns: 1.0000 ---------------------------------------------------- Final # 140 | TSS Error: 0.0925 | Correct: 1.0000 | RMS Error: 0.0439 Final # 140, Layer = 'output' | Units: 1.0000 | Patterns: 1.0000 ---------------------------------------------------- >>> test() Variations 1 2 3 4 5 6 7 8 j = 0 result = 1 j = 0 result = 1 j = 0 result = 1 j = 0 result = 1 j = 0 result = 1 j = 0 result = 1 j = 0 result = 1 j = 0 result = 1 j = 1 result = 2 j = 1 result = 2 j = 1 result = 4 X j = 1 result = 4 X j = 1 result = 4 X j = 1 result = 3 X j = 1 result = 2 j = 1 result = 2 j = 2 result = 3 j = 2 result = 3 j = 2 result = 3 j = 2 result = 4 X j = 2 result = 4 X j = 2 result = 3 j = 2 result = 4 X j = 2 result = 4 X j = 3 result = 4 j = 3 result = 4 j = 3 result = 4 j = 3 result = 4 j = 3 result = 4 j = 3 result = 4 j = 3 result = 4 j = 3 result = 4 #################################################################################################################### >>> train() Epoch # 10 | TSS Error: 9.4577 | Correct: 0.2917 | RMS Error: 0.4439 Epoch # 10, Layer = 'output' | Units: 0.2917 | Patterns: 0.0833 Epoch # 20 | TSS Error: 5.7488 | Correct: 0.2083 | RMS Error: 0.3461 Epoch # 20, Layer = 'output' | Units: 0.2083 | Patterns: 0.0000 Epoch # 30 | TSS Error: 8.4584 | Correct: 0.3958 | RMS Error: 0.4198 Epoch # 30, Layer = 'output' | Units: 0.3958 | Patterns: 0.0000 Epoch # 40 | TSS Error: 7.1423 | Correct: 0.5625 | RMS Error: 0.3857 Epoch # 40, Layer = 'output' | Units: 0.5625 | Patterns: 0.1667 Epoch # 50 | TSS Error: 5.6601 | Correct: 0.5000 | RMS Error: 0.3434 Epoch # 50, Layer = 'output' | Units: 0.5000 | Patterns: 0.0000 Epoch # 60 | TSS Error: 6.5893 | Correct: 0.6667 | RMS Error: 0.3705 Epoch # 60, Layer = 'output' | Units: 0.6667 | Patterns: 0.4167 Epoch # 70 | TSS Error: 7.9724 | Correct: 0.6250 | RMS Error: 0.4075 Epoch # 70, Layer = 'output' | Units: 0.6250 | Patterns: 0.2500 Epoch # 80 | TSS Error: 9.8965 | Correct: 0.5833 | RMS Error: 0.4541 Epoch # 80, Layer = 'output' | Units: 0.5833 | Patterns: 0.2500 Epoch # 90 | TSS Error: 7.7567 | Correct: 0.5833 | RMS Error: 0.4020 Epoch # 90, Layer = 'output' | Units: 0.5833 | Patterns: 0.2500 Epoch # 100 | TSS Error: 3.3880 | Correct: 0.8125 | RMS Error: 0.2657 Epoch # 100, Layer = 'output' | Units: 0.8125 | Patterns: 0.5000 Epoch # 110 | TSS Error: 4.2753 | Correct: 0.6875 | RMS Error: 0.2984 Epoch # 110, Layer = 'output' | Units: 0.6875 | Patterns: 0.5000 Epoch # 120 | TSS Error: 5.1072 | Correct: 0.7917 | RMS Error: 0.3262 Epoch # 120, Layer = 'output' | Units: 0.7917 | Patterns: 0.4167 Epoch # 130 | TSS Error: 3.8045 | Correct: 0.7917 | RMS Error: 0.2815 Epoch # 130, Layer = 'output' | Units: 0.7917 | Patterns: 0.5000 ---------------------------------------------------- Final # 136 | TSS Error: 0.0875 | Correct: 1.0000 | RMS Error: 0.0427 Final # 136, Layer = 'output' | Units: 1.0000 | Patterns: 1.0000 ---------------------------------------------------- >>> test() Variations 1 2 3 4 5 6 7 8 j = 0 result = 1 j = 0 result = 1 j = 0 result = 1 j = 0 result = 1 j = 0 result = 1 j = 0 result = 1 j = 0 result = 1 j = 0 result = 1 j = 1 result = 2 j = 1 result = 2 j = 1 result = 2 j = 1 result = 4 X j = 1 result = 4 X j = 1 result = 3 X j = 1 result = 2 j = 1 result = 2 j = 2 result = 3 j = 2 result = 3 j = 2 result = 3 j = 2 result = 4 X j = 2 result = 4 X j = 2 result = 3 j = 2 result = 4 X j = 2 result = 4 X j = 3 result = 4 j = 3 result = 4 j = 3 result = 4 j = 3 result = 4 j = 3 result = 4 j = 3 result = 4 j = 3 result = 4 j = 3 result = 4
Read a little more about networks as well.
5 June 2009 Expanded on yesterday's network, messing with various variables in order to try to get a network that could recognize no shape, circles, triangles, squares, trapezoids, and parallelograms ("shapes" group). Also tried getting the network to recognize shapes regardless of their position in the image ("blah" group"). Both could recognize the shapes each was trained on, but had trouble with the others (if small shapes were moved in the picture, for example, the network would guess that there wasn't a shape in the image):
>>> train() shape set = shapes # shapes = 4 tolerance = 0.3 hidden = 50 epsilon = 0.5 Epoch # 10 | TSS Error: 8.7614 | Correct: 0.6458 | RMS Error: 0.4272 Epoch # 10, Layer = 'output' | Units: 0.6458 | Patterns: 0.0000 Epoch # 20 | TSS Error: 7.7744 | Correct: 0.6458 | RMS Error: 0.4024 Epoch # 20, Layer = 'output' | Units: 0.6458 | Patterns: 0.1667 Epoch # 30 | TSS Error: 5.3608 | Correct: 0.7917 | RMS Error: 0.3342 Epoch # 30, Layer = 'output' | Units: 0.7917 | Patterns: 0.4167 Epoch # 40 | TSS Error: 5.5421 | Correct: 0.7083 | RMS Error: 0.3398 Epoch # 40, Layer = 'output' | Units: 0.7083 | Patterns: 0.3333 Epoch # 50 | TSS Error: 6.0068 | Correct: 0.8125 | RMS Error: 0.3538 Epoch # 50, Layer = 'output' | Units: 0.8125 | Patterns: 0.5000 Epoch # 60 | TSS Error: 4.4176 | Correct: 0.7292 | RMS Error: 0.3034 Epoch # 60, Layer = 'output' | Units: 0.7292 | Patterns: 0.4167 Epoch # 70 | TSS Error: 8.2014 | Correct: 0.6250 | RMS Error: 0.4134 Epoch # 70, Layer = 'output' | Units: 0.6250 | Patterns: 0.3333 Epoch # 80 | TSS Error: 7.5126 | Correct: 0.7083 | RMS Error: 0.3956 Epoch # 80, Layer = 'output' | Units: 0.7083 | Patterns: 0.3333 Epoch # 90 | TSS Error: 5.0271 | Correct: 0.6875 | RMS Error: 0.3236 Epoch # 90, Layer = 'output' | Units: 0.6875 | Patterns: 0.4167 Epoch # 100 | TSS Error: 6.3066 | Correct: 0.6875 | RMS Error: 0.3625 Epoch # 100, Layer = 'output' | Units: 0.6875 | Patterns: 0.2500 Epoch # 110 | TSS Error: 4.3986 | Correct: 0.7917 | RMS Error: 0.3027 Epoch # 110, Layer = 'output' | Units: 0.7917 | Patterns: 0.4167 Epoch # 120 | TSS Error: 4.1323 | Correct: 0.8125 | RMS Error: 0.2934 Epoch # 120, Layer = 'output' | Units: 0.8125 | Patterns: 0.4167 Epoch # 130 | TSS Error: 4.2382 | Correct: 0.8542 | RMS Error: 0.2971 Epoch # 130, Layer = 'output' | Units: 0.8542 | Patterns: 0.6667 Epoch # 140 | TSS Error: 2.6472 | Correct: 0.9167 | RMS Error: 0.2348 Epoch # 140, Layer = 'output' | Units: 0.9167 | Patterns: 0.7500 Epoch # 150 | TSS Error: 2.0734 | Correct: 0.8750 | RMS Error: 0.2078 Epoch # 150, Layer = 'output' | Units: 0.8750 | Patterns: 0.7500 Final # 157 | TSS Error: 0.5877 | Correct: 1.0000 | RMS Error: 0.1107 Final # 157, Layer = 'output' | Units: 1.0000 | Patterns: 1.0000 >>> test() Variations 0 1 2 3 4 5 6 7 8 Shape 0 0 0 0 0 0 0 0 0 0 Shape 1 1 1 1 3 X 3 X 2 X 3 X 1 2 X Shape 2 2 2 2 3 X 3 X 2 3 X 3 X 2 Shape 3 3 3 3 3 3 3 3 3 2 X >>> ================================ RESTART ================================ >>> Conx, version 2484 (psyco enabled) Conx using seed: 1244269222.52 >>> train() shape set = shapes # shapes = 5 tolerance = 0.3 hidden = 50 epsilon = 0.5 Epoch # 10 | TSS Error: 11.8153 | Correct: 0.7333 | RMS Error: 0.3969 Epoch # 10, Layer = 'output' | Units: 0.7333 | Patterns: 0.0667 Epoch # 20 | TSS Error: 8.0311 | Correct: 0.7867 | RMS Error: 0.3272 Epoch # 20, Layer = 'output' | Units: 0.7867 | Patterns: 0.3333 Epoch # 30 | TSS Error: 9.6049 | Correct: 0.7867 | RMS Error: 0.3579 Epoch # 30, Layer = 'output' | Units: 0.7867 | Patterns: 0.4667 Epoch # 40 | TSS Error: 7.5708 | Correct: 0.8400 | RMS Error: 0.3177 Epoch # 40, Layer = 'output' | Units: 0.8400 | Patterns: 0.5333 Epoch # 50 | TSS Error: 9.0984 | Correct: 0.7600 | RMS Error: 0.3483 Epoch # 50, Layer = 'output' | Units: 0.7600 | Patterns: 0.4000 Epoch # 60 | TSS Error: 7.3489 | Correct: 0.8133 | RMS Error: 0.3130 Epoch # 60, Layer = 'output' | Units: 0.8133 | Patterns: 0.4000 Epoch # 70 | TSS Error: 8.4983 | Correct: 0.8400 | RMS Error: 0.3366 Epoch # 70, Layer = 'output' | Units: 0.8400 | Patterns: 0.5333 Epoch # 80 | TSS Error: 4.7782 | Correct: 0.8533 | RMS Error: 0.2524 Epoch # 80, Layer = 'output' | Units: 0.8533 | Patterns: 0.6000 Epoch # 90 | TSS Error: 7.1936 | Correct: 0.8267 | RMS Error: 0.3097 Epoch # 90, Layer = 'output' | Units: 0.8267 | Patterns: 0.4667 Epoch # 100 | TSS Error: 9.2659 | Correct: 0.8533 | RMS Error: 0.3515 Epoch # 100, Layer = 'output' | Units: 0.8533 | Patterns: 0.6000 Epoch # 110 | TSS Error: 6.6857 | Correct: 0.8667 | RMS Error: 0.2986 Epoch # 110, Layer = 'output' | Units: 0.8667 | Patterns: 0.5333 Epoch # 120 | TSS Error: 4.1355 | Correct: 0.8267 | RMS Error: 0.2348 Epoch # 120, Layer = 'output' | Units: 0.8267 | Patterns: 0.6000 Epoch # 130 | TSS Error: 6.0246 | Correct: 0.8533 | RMS Error: 0.2834 Epoch # 130, Layer = 'output' | Units: 0.8533 | Patterns: 0.5333 Epoch # 140 | TSS Error: 0.8918 | Correct: 0.9333 | RMS Error: 0.1090 Epoch # 140, Layer = 'output' | Units: 0.9333 | Patterns: 0.6667 Epoch # 150 | TSS Error: 2.2225 | Correct: 0.9333 | RMS Error: 0.1721 Epoch # 150, Layer = 'output' | Units: 0.9333 | Patterns: 0.7333 Final # 153 | TSS Error: 0.4030 | Correct: 1.0000 | RMS Error: 0.0733 Final # 153, Layer = 'output' | Units: 1.0000 | Patterns: 1.0000 >>> test() Variations 0 1 2 3 4 5 6 7 8 Shape 0 0 0 0 0 0 0 0 0 0 Shape 1 1 1 1 3 X 3 X 2 X 1 1 2 X Shape 2 2 2 2 3 X 3 X 2 3 X 3 X 2 Shape 3 3 3 3 3 3 3 3 3 2 X Shape 4 4 4 2 X 3 X 3 X 2 X 4 4 2 X >>> ================================ RESTART ================================ >>> Conx, version 2484 (psyco enabled) Conx using seed: 1244321466.86 >>> train() shape set = blah # shapes = 5 tolerance = 0.3 hidden = 50 epsilon = 0.5 Epoch # 10 | TSS Error: 11.8399 | Correct: 0.6933 | RMS Error: 0.3973 Epoch # 10, Layer = 'output' | Units: 0.6933 | Patterns: 0.0000 Epoch # 20 | TSS Error: 9.3797 | Correct: 0.7067 | RMS Error: 0.3536 Epoch # 20, Layer = 'output' | Units: 0.7067 | Patterns: 0.2000 Epoch # 30 | TSS Error: 8.4218 | Correct: 0.8533 | RMS Error: 0.3351 Epoch # 30, Layer = 'output' | Units: 0.8533 | Patterns: 0.4000 Epoch # 40 | TSS Error: 5.2952 | Correct: 0.8800 | RMS Error: 0.2657 Epoch # 40, Layer = 'output' | Units: 0.8800 | Patterns: 0.4667 Epoch # 50 | TSS Error: 7.5040 | Correct: 0.8267 | RMS Error: 0.3163 Epoch # 50, Layer = 'output' | Units: 0.8267 | Patterns: 0.4000 Epoch # 60 | TSS Error: 4.3835 | Correct: 0.9067 | RMS Error: 0.2418 Epoch # 60, Layer = 'output' | Units: 0.9067 | Patterns: 0.6667 Epoch # 70 | TSS Error: 6.3656 | Correct: 0.8267 | RMS Error: 0.2913 Epoch # 70, Layer = 'output' | Units: 0.8267 | Patterns: 0.4667 Epoch # 80 | TSS Error: 4.8228 | Correct: 0.9067 | RMS Error: 0.2536 Epoch # 80, Layer = 'output' | Units: 0.9067 | Patterns: 0.7333 Epoch # 90 | TSS Error: 4.7728 | Correct: 0.8800 | RMS Error: 0.2523 Epoch # 90, Layer = 'output' | Units: 0.8800 | Patterns: 0.6667 Epoch # 100 | TSS Error: 5.2392 | Correct: 0.8533 | RMS Error: 0.2643 Epoch # 100, Layer = 'output' | Units: 0.8533 | Patterns: 0.6667 Epoch # 110 | TSS Error: 5.0018 | Correct: 0.8933 | RMS Error: 0.2582 Epoch # 110, Layer = 'output' | Units: 0.8933 | Patterns: 0.6667 Epoch # 120 | TSS Error: 2.7895 | Correct: 0.9333 | RMS Error: 0.1929 Epoch # 120, Layer = 'output' | Units: 0.9333 | Patterns: 0.8000 Epoch # 130 | TSS Error: 1.3352 | Correct: 0.9600 | RMS Error: 0.1334 Epoch # 130, Layer = 'output' | Units: 0.9600 | Patterns: 0.8667 Epoch # 140 | TSS Error: 5.2413 | Correct: 0.8667 | RMS Error: 0.2644 Epoch # 140, Layer = 'output' | Units: 0.8667 | Patterns: 0.5333 Epoch # 150 | TSS Error: 3.1829 | Correct: 0.9067 | RMS Error: 0.2060 Epoch # 150, Layer = 'output' | Units: 0.9067 | Patterns: 0.7333 Epoch # 160 | TSS Error: 4.2475 | Correct: 0.8533 | RMS Error: 0.2380 Epoch # 160, Layer = 'output' | Units: 0.8533 | Patterns: 0.6667 Epoch # 170 | TSS Error: 1.3375 | Correct: 0.9467 | RMS Error: 0.1335 Epoch # 170, Layer = 'output' | Units: 0.9467 | Patterns: 0.8667 Epoch # 180 | TSS Error: 3.5071 | Correct: 0.9067 | RMS Error: 0.2162 Epoch # 180, Layer = 'output' | Units: 0.9067 | Patterns: 0.7333 Epoch # 190 | TSS Error: 2.5185 | Correct: 0.9067 | RMS Error: 0.1832 Epoch # 190, Layer = 'output' | Units: 0.9067 | Patterns: 0.8000 Epoch # 200 | TSS Error: 2.3102 | Correct: 0.9067 | RMS Error: 0.1755 Epoch # 200, Layer = 'output' | Units: 0.9067 | Patterns: 0.8000 Epoch # 210 | TSS Error: 2.0388 | Correct: 0.9333 | RMS Error: 0.1649 Epoch # 210, Layer = 'output' | Units: 0.9333 | Patterns: 0.8667 Epoch # 220 | TSS Error: 1.2211 | Correct: 0.9600 | RMS Error: 0.1276 Epoch # 220, Layer = 'output' | Units: 0.9600 | Patterns: 0.8667 Epoch # 230 | TSS Error: 0.8683 | Correct: 0.9600 | RMS Error: 0.1076 Epoch # 230, Layer = 'output' | Units: 0.9600 | Patterns: 0.8667 Epoch # 240 | TSS Error: 1.4609 | Correct: 0.9733 | RMS Error: 0.1396 Epoch # 240, Layer = 'output' | Units: 0.9733 | Patterns: 0.9333 Final # 247 | TSS Error: 0.0179 | Correct: 1.0000 | RMS Error: 0.0155 Final # 247, Layer = 'output' | Units: 1.0000 | Patterns: 1.0000 >>> test() Variations 0 1 2 3 4 5 6 7 8 Shape 0 0 0 0 0 0 0 0 0 0 Shape 1 1 1 1 3 X 1 0 X 3 X 0 X 0 X Shape 2 2 2 2 3 X 3 X 0 X 3 X 0 X 0 X Shape 3 3 3 3 3 1 X 0 X 3 0 X 0 X Shape 4 4 4 4 1 X 0 X 0 X 2 X 0 X 0 X >>> ================================ RESTART ================================ >>> Conx, version 2484 (psyco enabled) Conx using seed: 1244317391.99 >>> train() shape set = blah # shapes = 5 tolerance = 0.3 hidden = 50 epsilon = 0.5 Epoch # 10 | TSS Error: 12.6897 | Correct: 0.6933 | RMS Error: 0.4113 Epoch # 10, Layer = 'output' | Units: 0.6933 | Patterns: 0.0000 Epoch # 20 | TSS Error: 7.0607 | Correct: 0.8667 | RMS Error: 0.3068 Epoch # 20, Layer = 'output' | Units: 0.8667 | Patterns: 0.5333 Epoch # 30 | TSS Error: 8.8573 | Correct: 0.7600 | RMS Error: 0.3437 Epoch # 30, Layer = 'output' | Units: 0.7600 | Patterns: 0.3333 Epoch # 40 | TSS Error: 4.2339 | Correct: 0.8400 | RMS Error: 0.2376 Epoch # 40, Layer = 'output' | Units: 0.8400 | Patterns: 0.6667 Epoch # 50 | TSS Error: 7.4077 | Correct: 0.8400 | RMS Error: 0.3143 Epoch # 50, Layer = 'output' | Units: 0.8400 | Patterns: 0.5333 Epoch # 60 | TSS Error: 4.1317 | Correct: 0.8933 | RMS Error: 0.2347 Epoch # 60, Layer = 'output' | Units: 0.8933 | Patterns: 0.6667 Epoch # 70 | TSS Error: 2.6337 | Correct: 0.9467 | RMS Error: 0.1874 Epoch # 70, Layer = 'output' | Units: 0.9467 | Patterns: 0.8000 Epoch # 80 | TSS Error: 4.8804 | Correct: 0.9200 | RMS Error: 0.2551 Epoch # 80, Layer = 'output' | Units: 0.9200 | Patterns: 0.7333 Epoch # 90 | TSS Error: 3.4393 | Correct: 0.9067 | RMS Error: 0.2141 Epoch # 90, Layer = 'output' | Units: 0.9067 | Patterns: 0.7333 Epoch # 100 | TSS Error: 3.4462 | Correct: 0.9067 | RMS Error: 0.2144 Epoch # 100, Layer = 'output' | Units: 0.9067 | Patterns: 0.7333 Epoch # 110 | TSS Error: 3.8030 | Correct: 0.8667 | RMS Error: 0.2252 Epoch # 110, Layer = 'output' | Units: 0.8667 | Patterns: 0.6000 Epoch # 120 | TSS Error: 1.4750 | Correct: 0.9600 | RMS Error: 0.1402 Epoch # 120, Layer = 'output' | Units: 0.9600 | Patterns: 0.8667 Epoch # 130 | TSS Error: 2.7820 | Correct: 0.9333 | RMS Error: 0.1926 Epoch # 130, Layer = 'output' | Units: 0.9333 | Patterns: 0.8000 Final # 138 | TSS Error: 0.3038 | Correct: 1.0000 | RMS Error: 0.0636 Final # 138, Layer = 'output' | Units: 1.0000 | Patterns: 1.0000 >>> test() Variations 0 1 2 3 4 5 6 7 8 Shape 0 2 X 2 X 2 X 2 X 2 X 2 X 2 X 2 X 2 X Shape 1 0 X 0 X 2 X 2 X 2 X 2 X 2 X 2 X 2 X Shape 2 1 X 1 X 1 X 2 2 2 2 2 2 Shape 3 2 X 2 X 2 X 2 X 2 X 2 X 2 X 3 2 X Shape 4 3 X 3 X 3 X 0 X 4 2 X 1 X 3 X 2 X >>> ================================ RESTART ================================ >>> Conx, version 2484 (psyco enabled) Conx using seed: 1244229005.5 >>> train() shape set = blah # shapes = 5 tolerance = 0.3 hidden = 50 epsilon = 0.5 Epoch # 10 | TSS Error: 12.4417 | Correct: 0.7333 | RMS Error: 0.4073 Epoch # 10, Layer = 'output' | Units: 0.7333 | Patterns: 0.0667 Epoch # 20 | TSS Error: 10.7405 | Correct: 0.7467 | RMS Error: 0.3784 Epoch # 20, Layer = 'output' | Units: 0.7467 | Patterns: 0.0000 Epoch # 30 | TSS Error: 5.0452 | Correct: 0.8933 | RMS Error: 0.2594 Epoch # 30, Layer = 'output' | Units: 0.8933 | Patterns: 0.5333 Epoch # 40 | TSS Error: 8.5339 | Correct: 0.8667 | RMS Error: 0.3373 Epoch # 40, Layer = 'output' | Units: 0.8667 | Patterns: 0.3333 Epoch # 50 | TSS Error: 3.3184 | Correct: 0.9467 | RMS Error: 0.2103 Epoch # 50, Layer = 'output' | Units: 0.9467 | Patterns: 0.7333 Epoch # 60 | TSS Error: 4.9180 | Correct: 0.9200 | RMS Error: 0.2561 Epoch # 60, Layer = 'output' | Units: 0.9200 | Patterns: 0.6000 Epoch # 70 | TSS Error: 7.1104 | Correct: 0.8667 | RMS Error: 0.3079 Epoch # 70, Layer = 'output' | Units: 0.8667 | Patterns: 0.5333 Epoch # 80 | TSS Error: 3.2825 | Correct: 0.9467 | RMS Error: 0.2092 Epoch # 80, Layer = 'output' | Units: 0.9467 | Patterns: 0.7333 Epoch # 90 | TSS Error: 2.1907 | Correct: 0.9600 | RMS Error: 0.1709 Epoch # 90, Layer = 'output' | Units: 0.9600 | Patterns: 0.8000 Epoch # 100 | TSS Error: 2.5514 | Correct: 0.9333 | RMS Error: 0.1844 Epoch # 100, Layer = 'output' | Units: 0.9333 | Patterns: 0.8000 Epoch # 110 | TSS Error: 2.5748 | Correct: 0.9333 | RMS Error: 0.1853 Epoch # 110, Layer = 'output' | Units: 0.9333 | Patterns: 0.8000 Epoch # 120 | TSS Error: 2.5978 | Correct: 0.9467 | RMS Error: 0.1861 Epoch # 120, Layer = 'output' | Units: 0.9467 | Patterns: 0.8000 Epoch # 130 | TSS Error: 2.6229 | Correct: 0.9333 | RMS Error: 0.1870 Epoch # 130, Layer = 'output' | Units: 0.9333 | Patterns: 0.8000 Epoch # 140 | TSS Error: 3.7036 | Correct: 0.9200 | RMS Error: 0.2222 Epoch # 140, Layer = 'output' | Units: 0.9200 | Patterns: 0.7333 Epoch # 150 | TSS Error: 2.3653 | Correct: 0.9333 | RMS Error: 0.1776 Epoch # 150, Layer = 'output' | Units: 0.9333 | Patterns: 0.8000 Epoch # 160 | TSS Error: 2.7573 | Correct: 0.9467 | RMS Error: 0.1917 Epoch # 160, Layer = 'output' | Units: 0.9467 | Patterns: 0.8667 Final # 166 | TSS Error: 0.0244 | Correct: 1.0000 | RMS Error: 0.0180 Final # 166, Layer = 'output' | Units: 1.0000 | Patterns: 1.0000 >>> test() Variations 0 1 2 3 4 5 6 7 8 Shape 0 0 0 0 0 0 0 0 0 0 Shape 1 1 1 1 1 1 0 X 1 0 X 0 X Shape 2 2 2 2 1 X 2 0 X 1 X 0 X 0 X Shape 3 3 3 3 1 X 0 X 0 X 1 X 0 X 0 X Shape 4 4 4 4 1 X 0 X 0 X 2 X 2 X 0 X
