Posts tagged with “AI”
adaptive training #3
it's amazing what changing the targets can do. now they're spread evenly from the starting position (uniform distance) and they seem to do much better, rather than finding a local maximum (a single target that's easier to get to than the others).
0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
avg | 0.1299111 | 0.1520748 | 0.2072772 | 0.3187084 | 0.3958412 | 0.6468749 | 0.6010848 | 0.7655407 | 0.9283564 | 0.9560807 | 1.27025 | 1.240294 | 1.078598 | 1.257212 | 1.092487 | 1.327467 | 1.08835 | 1.154903 | 1.370362 | 1.198214 | 1.211951 | 1.178578 | 1.253744 | 0.9513875 | 1.347026 | 0.8872428 |
max | 0.6187887 | 0.7656475 | 1.199693 | 1.060628 | 1.624553 | 2.014889 | 2.00272 | 1.532919 | 1.779993 | 2.924508 | 2.019435 | 1.890425 | 2.37094 | 2.055064 | 1.812547 | 2.125103 | 2.044051 | 1.91144 | 2.050521 | 2.57358 | 1.498748 | 1.417337 | 1.908807 | 1.261683 | 1.80807 | 1.39744 |
- Population Size
- 100
- Run Length
- 20
- Do Training
- False
- Do Hybrid Training
- True
- Do Competitive Run
- False
- Do Adaptive Training
- True
- Inputs
- hasTarget,targetLeft,targetRight,targetNorth,targetSouth,dirToTarget,distToTarget,wall,sensor0,sensor1,sensor2
- Outputs
- left,right,up,down,run,sensordir0,sensordir1,sensordir2
- Hidden Layers
- 8 12 12 12 8
- Back Propogation
- True
- Learning rate
- 0.5
- Momentum
- 0.1
- Growth rate
- 0.5
adaptive training #2
0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
avg | 0.11333 | 0.1216922 | 0.1410408 | 0.2255169 | 0.3438947 | 0.5308622 | 0.7092866 | 0.7473413 | 0.8380993 | 0.8315367 | 0.9242428 | 0.8393147 | 1.103198 | 1.054069 | 1.111683 | 1.051629 | 1.028054 | 1.043636 | 0.9677835 | 1.01511 | 1.045051 | 1.138088 | 1.002189 | 0.8751988 | 0.9423944 | 0.9026732 | 1.062976 | 0.9172046 | 1.334726 | 1.079931 | 0.7032812 | 1.377699 | 1.236436 | 1.741362 | 1.081763 | 1.341366 | 0.8942567 | 0.9266959 | 1.079842 | 1.067705 | 0.7557345 | 1.354618 | 0.9338629 | 1.070021 | 1.288732 | 1.196118 | 1.037447 |
max | 0.4842224 | 0.7318087 | 0.7872303 | 1.314361 | 1.167771 | 1.026197 | 1.422564 | 1.538746 | 1.474501 | 1.283453 | 1.423418 | 1.349116 | 1.614929 | 1.392465 | 1.4685 | 1.274314 | 1.203464 | 1.228827 | 1.08125 | 1.171278 | 1.106939 | 1.359823 | 1.106474 | 0.9218987 | 1.06358 | 1.007161 | 1.376667 | 0.9957669 | 1.447871 | 1.173347 | 0.8280649 | 1.515407 | 1.369468 | 1.839345 | 1.161362 | 1.377432 | 0.9441369 | 0.9500773 | 1.109659 | 1.092207 | 0.7708632 | 1.38819 | 0.9459927 | 1.093874 | 1.322104 | 1.24041 | 1.065003 |
- Population Size
- 100
- Run Length
- 20
- Do Training
- False
- Do Hybrid Training
- True
- Do Competitive Run
- False
- Do Adaptive Training
- True
- Inputs
- hasTarget,targetLeft,targetRight,targetNorth,targetSouth,dirToTarget,distToTarget,wall,sensor0,sensor1,sensor2
- Outputs
- left,right,up,down,run,sensordir0,sensordir1,sensordir2
- Hidden Layers
- 8 12 12 12 8
- Back Propogation
- True
- Learning rate
- 0.5
- Momentum
- 0.1
- Growth rate
- 0.5
adaptive training
0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
avg | 0.1054728 | 0.1185876 | 0.1258131 | 0.130587 | 0.1506598 | 0.2122319 | 0.4581497 | 0.8586704 | 1.20199 | 1.170219 | 0.9763492 | 1.176481 | 0.938668 | 0.9572123 | 0.902207 | 1.133812 | 1.098755 | 0.9892185 | 1.5469 | 1.384855 | 1.156185 | 0.8816531 | 1.155313 | 1.370298 | 0.9086894 | 1.091801 | 0.9211531 | 1.064359 | 1.445552 | 0.9272204 | 1.219871 | 1.281623 | 1.345069 | 1.290509 |
max | 0.1787508 | 0.3422044 | 0.4880244 | 0.4869769 | 1.047815 | 1.288545 | 1.504831 | 1.27506 | 1.43819 | 1.796796 | 1.367374 | 1.229414 | 1.067725 | 1.265468 | 1.230431 | 1.216018 | 1.179089 | 1.173458 | 1.681361 | 1.496624 | 1.225676 | 1.205523 | 1.230803 | 1.436737 | 1.25253 | 1.174299 | 0.9801109 | 1.207027 | 1.711045 | 1.16089 | 1.435449 | 1.345895 | 1.40991 | 1.366372 |
- Population Size
- 100
- Run Length
- 20
- Do Training
- False
- Do Hybrid Training
- True
- Do Competitive Run
- False
- Do Adaptive Training
- True
- Inputs
- hasTarget,targetLeft,targetRight,targetNorth,targetSouth,dirToTarget,distToTarget,wall,sensor0,sensor1,sensor2
- Outputs
- left,right,up,down,run,sensordir0,sensordir1,sensordir2
- Hidden Layers
- 8 12 12 12 8
- Back Propogation
- True
- Learning rate
- 0.5
- Momentum
- 0.1
- Growth rate
- 0.5
overhead run #4
and a run w/another hidden layer
0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
avg | 0.1104814 | 0.136816 | 0.1908503 | 0.3284849 | 0.6684353 | 1.282734 | 1.302261 | 1.206301 | 1.358841 | 1.126256 | 1.397736 | 1.246604 | 1.323351 | 0.8909027 | 1.122478 | 1.108233 | 1.631226 | 1.291764 | 1.220225 | 1.21547 | 1.299522 | 1.295794 | 1.360445 | 1.006665 | 1.005265 | 1.285506 | 1.723688 | 0.9848436 | 0.9455433 | 0.9313265 | 1.148989 | 1.272345 |
max | 0.2871341 | 1.214576 | 1.067269 | 1.103958 | 1.354696 | 1.475417 | 1.366202 | 1.234541 | 1.390623 | 1.173298 | 1.426518 | 1.764977 | 1.396152 | 0.9108025 | 1.144939 | 1.133509 | 1.90503 | 1.320514 | 1.246103 | 1.239381 | 1.498761 | 1.339491 | 1.46306 | 1.135313 | 1.030648 | 1.328566 | 1.772488 | 1.166132 | 0.9746216 | 1.090431 | 1.179814 | 1.307164 |
- Population Size
- 100
- Run Length
- 20
- Do Training
- False
- Do Hybrid Training
- True
- Do Competitive Run
- False
- Inputs
- hasTarget,targetLeft,targetRight,targetNorth,targetSouth,dirToTarget,distToTarget,wall,sensor0,sensor1,sensor2
- Outputs
- left,right,up,down,run,sensordir0,sensordir1,sensordir2
- Hidden Layers
- 8 12 12 12 8
- Back Propogation
- True
- Learning rate
- 0.5
- Momentum
- 0.1
- Growth rate
- 0.5