Abstract: This paper discusses about KNN algorithm and its various modified versions available in formerly done studies. Despite the simplicity of KNN algorithm there are a number of shortcomings in it i.e. high computational cost, large memory requirement, equal-weighted features and in last deciding appropriate value of the input parameter k. Many researchers have proposed modified versions of KNN algorithm to overcome these shortcomings. In this study, we will analyze profoundly these variants of KNN algorithm and their performances.

Keywords: Evolutionary Computing; KNN; lazy learner; metaheuristic; nearest neighbour.