Nearest Neighbor Ensembles Combines with Weighted Instance and Feature Sub Set Selection: A Survey

Ensemble learning deals with methods which employ multiple learners to solve a problem The generalization ability of an ensemble is usually significantly better than that of a single learner, so ensemble methods are very attractive, at the same time feature selection process of ensemble technique has important role of classifier. This paper, presents the analysis on classification technique of k-nearest neighbor method while applying feature reduction in subsets and further more constructing ensembles of nearest neighbor classifiers on the basis of the instance selection. Instance selection is to obtain the subset of the instances available for training capable of achieving, at least, the same performance as the whole training set will provide. Keywords— Ensemble learning, k-nearest neighbor, feature sub set selection and instance selection.