Exclusive lasso-based k-nearest-neighbor classification

Conventionally, the k nearest-neighbor (kNN) classification is implemented with use of Euclidean distance-based measures, which are mainly one-to-one similarity relationships such as to lose connections between different samples. As a strategy alleviate this issue, coefficients coded by sparse representation have played role gauger for well. Although SR enjoy remarkable discrimination nature one-to-many relationship, it carries out variable selection at individual level so that possible inherent group structure ignored. In order make most information implied in structure, paper employs exclusive lasso perform evaluation two novel methods. Experimental results on both benchmark data sets and face recognition problem demonstrate EL-based kNN method outperforms certain state-of-the-art techniques existing representation-based approaches, terms size feature reduction accuracy.