A Novel Approach for Feature Selection and Classification of Diabetes Mellitus: Machine Learning Methods

Table 3

Accuracy of classifiers for different feature selection technique.

N	Algorithm	Correlation attribute	Information	Gain	Principal component
6	Multilayer perceptron	75.1	74.8		74.0
	Decision trees	74.3	74.2		73.5
	Random forest	74.2	74.6		75.1
	K-nearest neighbour	67.0	68.0		65.7
4	Multilayer perceptron	75.1	76.9		72.6
	Decision trees	74.0	74.3		72.5
	Random forest	73.3	71.7		72.3
	K-nearest neighbour	70.1	68.0		65.7