CL-USER 9 > (load "452/lisp/backprop") ; Loading fasl file 452/lisp/backprop.wfasl #P"/u2/instructors/paulmac/452/lisp/backprop.wfasl" CL-USER 10 > (build-network :initial-weights 0.2) NIL CL-USER 11 > (train-network :iterations 500) After 0 iterations, the network achieved the following on 4 test patterns. Sum Squared Error: 0.50001 Percent Correct: 50.000 (with a threshold of 0.50) weights: (To: HID-1 Val: 0.000 To: HID-2 Val: 0.000 To: HID-3 Val: 0.000 To: HID-4 Val: 0.000) weights: (To: HID-1 Val: 0.096 To: HID-2 Val: 0.166 To: HID-3 Val: -0.076 To: HID-4 Val: -0.179) weights: (To: HID-1 Val: 0.065 To: HID-2 Val: -0.075 To: HID-3 Val: 0.084 To: HID-4 Val: 0.027) weights: (To: OUT-1 Val: 0.000) weights: (To: OUT-1 Val: 0.072) weights: (To: OUT-1 Val: 0.050) weights: (To: OUT-1 Val: 0.002) weights: (To: OUT-1 Val: -0.121) After 100 iterations, the network achieved the following on 4 test patterns. Sum Squared Error: 0.50010 Percent Correct: 50.000 (with a threshold of 0.50) weights: (To: HID-1 Val: -0.000 To: HID-2 Val: -0.003 To: HID-3 Val: -0.000 To: HID-4 Val: 0.003) weights: (To: HID-1 Val: 0.095 To: HID-2 Val: 0.162 To: HID-3 Val: -0.076 To: HID-4 Val: -0.173) weights: (To: HID-1 Val: 0.067 To: HID-2 Val: -0.076 To: HID-3 Val: 0.084 To: HID-4 Val: 0.023) weights: (To: OUT-1 Val: -0.017) weights: (To: OUT-1 Val: 0.063) weights: (To: OUT-1 Val: 0.040) weights: (To: OUT-1 Val: -0.006) weights: (To: OUT-1 Val: -0.127) After 200 iterations, the network achieved the following on 4 test patterns. Sum Squared Error: 0.50793 Percent Correct: 50.000 (with a threshold of 0.50) weights: (To: HID-1 Val: -0.029 To: HID-2 Val: -0.032 To: HID-3 Val: -0.022 To: HID-4 Val: -0.003) weights: (To: HID-1 Val: 0.080 To: HID-2 Val: 0.145 To: HID-3 Val: -0.087 To: HID-4 Val: -0.172) weights: (To: HID-1 Val: 0.054 To: HID-2 Val: -0.089 To: HID-3 Val: 0.073 To: HID-4 Val: 0.014) weights: (To: OUT-1 Val: -0.130) weights: (To: OUT-1 Val: 0.003) weights: (To: OUT-1 Val: -0.019) weights: (To: OUT-1 Val: -0.061) weights: (To: OUT-1 Val: -0.177) After 300 iterations, the network achieved the following on 4 test patterns. Sum Squared Error: 0.53493 Percent Correct: 50.000 (with a threshold of 0.50) weights: (To: HID-1 Val: -0.527 To: HID-2 Val: -0.527 To: HID-3 Val: -0.524 To: HID-4 Val: -0.535) weights: (To: HID-1 Val: -0.167 To: HID-2 Val: -0.100 To: HID-3 Val: -0.332 To: HID-4 Val: -0.435) weights: (To: HID-1 Val: -0.191 To: HID-2 Val: -0.332 To: HID-3 Val: -0.177 To: HID-4 Val: -0.275) weights: (To: OUT-1 Val: 0.367) weights: (To: OUT-1 Val: 0.182) weights: (To: OUT-1 Val: 0.165) weights: (To: OUT-1 Val: 0.136) weights: (To: OUT-1 Val: 0.052) After 400 iterations, the network achieved the following on 4 test patterns. Sum Squared Error: 0.51252 Percent Correct: 50.000 (with a threshold of 0.50) weights: (To: HID-1 Val: -0.953 To: HID-2 Val: -0.935 To: HID-3 Val: -0.909 To: HID-4 Val: -0.857) weights: (To: HID-1 Val: -0.342 To: HID-2 Val: -0.270 To: HID-3 Val: -0.510 To: HID-4 Val: -0.667) weights: (To: HID-1 Val: -0.368 To: HID-2 Val: -0.505 To: HID-3 Val: -0.365 To: HID-4 Val: -0.548) weights: (To: OUT-1 Val: -0.235) weights: (To: OUT-1 Val: -0.069) weights: (To: OUT-1 Val: -0.080) weights: (To: OUT-1 Val: -0.116) weights: (To: OUT-1 Val: -0.225) After 500 iterations, the network achieved the following on 4 test patterns. Sum Squared Error: 0.49527 Percent Correct: 50.000 (with a threshold of 0.50) weights: (To: HID-1 Val: -1.020 To: HID-2 Val: -1.006 To: HID-3 Val: -0.989 To: HID-4 Val: -0.867) weights: (To: HID-1 Val: -0.408 To: HID-2 Val: -0.373 To: HID-3 Val: -0.658 To: HID-4 Val: -1.171) weights: (To: HID-1 Val: -0.431 To: HID-2 Val: -0.573 To: HID-3 Val: -0.553 To: HID-4 Val: -1.121) weights: (To: OUT-1 Val: 0.245) weights: (To: OUT-1 Val: -0.046) weights: (To: OUT-1 Val: -0.069) weights: (To: OUT-1 Val: -0.156) weights: (To: OUT-1 Val: -0.491) NIL CL-USER 12 > (train-network :iterations 500) After 501 iterations, the network achieved the following on 4 test patterns. Sum Squared Error: 0.49247 Percent Correct: 75.000 (with a threshold of 0.50) weights: (To: HID-1 Val: -1.021 To: HID-2 Val: -1.007 To: HID-3 Val: -0.987 To: HID-4 Val: -0.822) weights: (To: HID-1 Val: -0.415 To: HID-2 Val: -0.385 To: HID-3 Val: -0.677 To: HID-4 Val: -1.257) weights: (To: HID-1 Val: -0.437 To: HID-2 Val: -0.580 To: HID-3 Val: -0.578 To: HID-4 Val: -1.214) weights: (To: OUT-1 Val: 0.223) weights: (To: OUT-1 Val: -0.058) weights: (To: OUT-1 Val: -0.082) weights: (To: OUT-1 Val: -0.178) weights: (To: OUT-1 Val: -0.556) After 601 iterations, the network achieved the following on 4 test patterns. Sum Squared Error: 0.43158 Percent Correct: 75.000 (with a threshold of 0.50) weights: (To: HID-1 Val: -1.038 To: HID-2 Val: -1.029 To: HID-3 Val: -0.995 To: HID-4 Val: 0.380) weights: (To: HID-1 Val: -0.477 To: HID-2 Val: -0.514 To: HID-3 Val: -0.973 To: HID-4 Val: -2.837) weights: (To: HID-1 Val: -0.496 To: HID-2 Val: -0.660 To: HID-3 Val: -0.921 To: HID-4 Val: -2.832) weights: (To: OUT-1 Val: 0.370) weights: (To: OUT-1 Val: -0.051) weights: (To: OUT-1 Val: -0.100) weights: (To: OUT-1 Val: -0.318) weights: (To: OUT-1 Val: -1.592) After 701 iterations, the network achieved the following on 4 test patterns. Sum Squared Error: 0.35272 Percent Correct: 75.000 (with a threshold of 0.50) weights: (To: HID-1 Val: -1.005 To: HID-2 Val: -1.014 To: HID-3 Val: -1.108 To: HID-4 Val: 1.577) weights: (To: HID-1 Val: -0.528 To: HID-2 Val: -0.543 To: HID-3 Val: -1.222 To: HID-4 Val: -5.233) weights: (To: HID-1 Val: -0.549 To: HID-2 Val: -0.691 To: HID-3 Val: -1.192 To: HID-4 Val: -5.234) weights: (To: OUT-1 Val: 0.581) weights: (To: OUT-1 Val: 0.289) weights: (To: OUT-1 Val: 0.242) weights: (To: OUT-1 Val: -0.121) weights: (To: OUT-1 Val: -2.883) After 801 iterations, the network achieved the following on 4 test patterns. Sum Squared Error: 0.31927 Percent Correct: 75.000 (with a threshold of 0.50) weights: (To: HID-1 Val: -0.506 To: HID-2 Val: -0.534 To: HID-3 Val: -1.078 To: HID-4 Val: 2.110) weights: (To: HID-1 Val: -0.771 To: HID-2 Val: -0.630 To: HID-3 Val: -1.212 To: HID-4 Val: -6.444) weights: (To: HID-1 Val: -0.798 To: HID-2 Val: -0.938 To: HID-3 Val: -1.180 To: HID-4 Val: -6.452) weights: (To: OUT-1 Val: 0.359) weights: (To: OUT-1 Val: 0.824) weights: (To: OUT-1 Val: 0.784) weights: (To: OUT-1 Val: 0.251) weights: (To: OUT-1 Val: -3.464) After 901 iterations, the network achieved the following on 4 test patterns. Sum Squared Error: 0.21083 Percent Correct: 100.000 (with a threshold of 0.50) weights: (To: HID-1 Val: 1.501 To: HID-2 Val: 1.230 To: HID-3 Val: -0.697 To: HID-4 Val: 2.698) weights: (To: HID-1 Val: -1.580 To: HID-2 Val: -0.952 To: HID-3 Val: -1.051 To: HID-4 Val: -7.218) weights: (To: HID-1 Val: -1.216 To: HID-2 Val: -1.680 To: HID-3 Val: -0.919 To: HID-4 Val: -7.265) weights: (To: OUT-1 Val: -0.718) weights: (To: OUT-1 Val: 1.670) weights: (To: OUT-1 Val: 1.574) weights: (To: OUT-1 Val: 0.615) weights: (To: OUT-1 Val: -4.086) After 1001 iterations, the network achieved the following on 4 test patterns. Sum Squared Error: 0.05587 Percent Correct: 100.000 (with a threshold of 0.50) weights: (To: HID-1 Val: 4.595 To: HID-2 Val: 4.062 To: HID-3 Val: -0.463 To: HID-4 Val: 3.268) weights: (To: HID-1 Val: -3.022 To: HID-2 Val: -2.600 To: HID-3 Val: -0.959 To: HID-4 Val: -7.837) weights: (To: HID-1 Val: -2.963 To: HID-2 Val: -2.720 To: HID-3 Val: -0.610 To: HID-4 Val: -7.874) weights: (To: OUT-1 Val: -2.209) weights: (To: OUT-1 Val: 2.642) weights: (To: OUT-1 Val: 2.355) weights: (To: OUT-1 Val: 0.680) weights: (To: OUT-1 Val: -5.021) NIL CL-USER 13 > (test-network) For input: ((0 0)), the output vector is [(0.0883473)] For input: ((0 1)), the output vector is [(0.7558313)] For input: ((1 0)), the output vector is [(0.753708)] For input: ((1 1)), the output vector is [(0.122715734)] NIL CL-USER 14 > (train-network :iterations 500) After 1002 iterations, the network achieved the following on 4 test patterns. Sum Squared Error: 0.07157 Percent Correct: 100.000 (with a threshold of 0.50) weights: (To: HID-1 Val: 4.411 To: HID-2 Val: 3.903 To: HID-3 Val: -0.537 To: HID-4 Val: 3.344) weights: (To: HID-1 Val: -3.444 To: HID-2 Val: -3.007 To: HID-3 Val: -0.992 To: HID-4 Val: -7.875) weights: (To: HID-1 Val: -3.404 To: HID-2 Val: -3.109 To: HID-3 Val: -0.628 To: HID-4 Val: -7.910) weights: (To: OUT-1 Val: -2.471) weights: (To: OUT-1 Val: 2.608) weights: (To: OUT-1 Val: 2.310) weights: (To: OUT-1 Val: 0.640) weights: (To: OUT-1 Val: -5.115) After 1102 iterations, the network achieved the following on 4 test patterns. Sum Squared Error: 0.50144 Percent Correct: 75.000 (with a threshold of 0.50) weights: (To: HID-1 Val: 10.040 To: HID-2 Val: 8.787 To: HID-3 Val: -1.422 To: HID-4 Val: 3.238) weights: (To: HID-1 Val: -4.267 To: HID-2 Val: -3.961 To: HID-3 Val: -1.564 To: HID-4 Val: -8.602) weights: (To: HID-1 Val: -4.246 To: HID-2 Val: -4.042 To: HID-3 Val: -0.981 To: HID-4 Val: -8.619) weights: (To: OUT-1 Val: -1.319) weights: (To: OUT-1 Val: 3.361) weights: (To: OUT-1 Val: 2.777) weights: (To: OUT-1 Val: 0.557) weights: (To: OUT-1 Val: -6.193) After 1202 iterations, the network achieved the following on 4 test patterns. Sum Squared Error: 0.01103 Percent Correct: 100.000 (with a threshold of 0.50) weights: (To: HID-1 Val: 10.453 To: HID-2 Val: 8.785 To: HID-3 Val: -1.797 To: HID-4 Val: 3.922) weights: (To: HID-1 Val: -7.937 To: HID-2 Val: -6.457 To: HID-3 Val: -1.601 To: HID-4 Val: -8.952) weights: (To: HID-1 Val: -7.915 To: HID-2 Val: -6.511 To: HID-3 Val: -0.924 To: HID-4 Val: -8.963) weights: (To: OUT-1 Val: -1.938) weights: (To: OUT-1 Val: 2.849) weights: (To: OUT-1 Val: 3.195) weights: (To: OUT-1 Val: 0.484) weights: (To: OUT-1 Val: -7.200) After 1302 iterations, the network achieved the following on 4 test patterns. Sum Squared Error: 0.00523 Percent Correct: 100.000 (with a threshold of 0.50) weights: (To: HID-1 Val: 10.954 To: HID-2 Val: 9.289 To: HID-3 Val: -1.883 To: HID-4 Val: 4.025) weights: (To: HID-1 Val: -7.742 To: HID-2 Val: -6.421 To: HID-3 Val: -1.609 To: HID-4 Val: -9.054) weights: (To: HID-1 Val: -7.727 To: HID-2 Val: -6.460 To: HID-3 Val: -0.908 To: HID-4 Val: -9.063) weights: (To: OUT-1 Val: -2.523) weights: (To: OUT-1 Val: 2.879) weights: (To: OUT-1 Val: 3.212) weights: (To: OUT-1 Val: 0.468) weights: (To: OUT-1 Val: -7.300) After 1402 iterations, the network achieved the following on 4 test patterns. Sum Squared Error: 0.00407 Percent Correct: 100.000 (with a threshold of 0.50) weights: (To: HID-1 Val: 11.206 To: HID-2 Val: 9.557 To: HID-3 Val: -1.910 To: HID-4 Val: 4.031) weights: (To: HID-1 Val: -7.699 To: HID-2 Val: -6.499 To: HID-3 Val: -1.606 To: HID-4 Val: -9.155) weights: (To: HID-1 Val: -7.688 To: HID-2 Val: -6.530 To: HID-3 Val: -0.878 To: HID-4 Val: -9.162) weights: (To: OUT-1 Val: -2.741) weights: (To: OUT-1 Val: 2.983) weights: (To: OUT-1 Val: 3.308) weights: (To: OUT-1 Val: 0.465) weights: (To: OUT-1 Val: -7.341) After 1502 iterations, the network achieved the following on 4 test patterns. Sum Squared Error: 0.00339 Percent Correct: 100.000 (with a threshold of 0.50) weights: (To: HID-1 Val: 11.357 To: HID-2 Val: 9.756 To: HID-3 Val: -1.934 To: HID-4 Val: 4.053) weights: (To: HID-1 Val: -7.716 To: HID-2 Val: -6.598 To: HID-3 Val: -1.601 To: HID-4 Val: -9.244) weights: (To: HID-1 Val: -7.708 To: HID-2 Val: -6.624 To: HID-3 Val: -0.847 To: HID-4 Val: -9.250) weights: (To: OUT-1 Val: -2.885) weights: (To: OUT-1 Val: 3.077) weights: (To: OUT-1 Val: 3.396) weights: (To: OUT-1 Val: 0.463) weights: (To: OUT-1 Val: -7.381) NIL CL-USER 15 > (test-network) For input: ((0 0)), the output vector is [(0.026403267)] For input: ((0 1)), the output vector is [(0.9664646)] For input: ((1 0)), the output vector is [(0.96609616)] For input: ((1 1)), the output vector is [(0.060893092)] NIL CL-USER 16 > (dribble)