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