Re-Test XOR - Githubissues

Amitai5 / Backprop-VS-Genetics

The GitHub Repository For The Research Paper "Supervised Data in Backpropagation and Genetics Learning Algorithms"

MIT License

1 stars 0 forks source link

Re-Test XOR #11

Closed Amitai5 closed 5 years ago

Amitai5 commented 5 years ago

The XOR test from before while it still was technically valid, the starting position for the neural networks was not a normal distribution. This puts Backpropagation at a disadvantage making it harder for it to find a global minimum.

Amitai5 commented 5 years ago

Backpropagation Algorithm Results

	Epoch Count	Time (Milliseconds)
Minimum:	2819	163.9362
1st Quartile:	4150.5	239.348025
3rd Quartile:	7475.5	406.024125
Maximum:	12543	647.6394

Amitai5 commented 5 years ago

Genetics Algorithm Results

	Generation Count	Time (Milliseconds)
Minimum:	101	169.9704
1st Quartile:	122.25	214.5134
3rd Quartile:	222	335.4027
Maximum:	1435	1863.1769