Network structure:			(8 4 8)
Epochs:					100
Test after every N epochs:		10
Learning rate (eta):			0.35
Momentum (alpha):			0.90
Noise?:					0
Training data:				politics.train
Testing data:				politics.test