A Prediction Model for Membrane Proteins Using Moments Based Features

Table 3

Confusion matrices of the neural networks in membrane protein classification.

	Target class
	Total outcome	Condition, positive	Condition, negative	Total percentage
Output class	Test outcome, positive	6824	605	91.85%
		43.89%	3.89%	8.15%
	Test outcome, negative	758	7360	90.66%
		4.87%	47.34%	9.34%
	Output accuracy	90.0%	92.4%	91.23%
		10.0%	7.6%	8.77%