KNN (k nearest neighbors) algorithm was proposed by Evelyn Fix and Joseph Hodges in 1951 and its theoretical system was moulded in (Cover & Hart 1967). Due to its simplicity, efficiency, easy-implementation, non-parametric, KNN classification was selected as one of top 10 algorithms in data mining and machine learning (Wu, et al. 2008). From human’s base action in everyday think, do and learn, KNN algorithm is actually the most natural solution. It is undoubted that KNN algorithm must be one of the most hopeful machine learning methods in artificial intelligence.However, KNN algorithm is a lazy learning procedure that has become the bottleneck constraining its widely applications. Apart from this, there are still some other challenges in KNN classification applications (Zhang 2022). To make it enter our life widely, we discuss the power of KNN algorithm and present some strategies of fighting for some challenges in this paper.