An Enhancement of K-Nearest Neighbor Algorithm Using Information Gain and Extension Relativity

An enhanced K-NN algorithm is proposed in this paper to improve the conventional K-NN algorithm. The enhanced K-NN algorithm proposed uses information gain and extension relativity. The weight coefficient is got through computing the information gain of attributes. In this approach, the anti-jamming ability and accuracy of the K-NN algorithm is improved highly, and the computing time is reduced and the time is improved greatly. The test results show that the novel K-NN algorithm is feasible and effective.