k*-Nearest Neighbors: From Global to Local

Abstract

The weighted k-nearest neighbors algorithm is one of the most fundamental non-parametric methods in pattern recognition and machine learning. The question of setting the optimal number of neighbors as well as the optimal weights has received much attention throughout the years, nevertheless this problem seems to have remained unsettled. In this paper we offer a simple approach to locally weighted regression/classification, where we make the bias-variance tradeoff explicit. Our formulation enables us to phrase a notion of optimal weights, and to efficiently find these weights as well as the optimal number of neighbors efficiently and adaptively, for each data point whose value we wish to estimate. The applicability of our approach is demonstrated on several datasets, showing superior performance over standard locally weighted methods.

Cited by in corpus (5)

• Nearest Neighbor and Kernel Survival Analysis: Nonasymptotic Error Bounds and Strong Consistency Rates
• Chest X-Rays Image Classification from beta-Variational Autoencoders Latent Features
• Towards Personalized and Human-in-the-Loop Document Summarization
• Efficient Balanced Treatment Assignments for Experimentation
• Extrapolation Towards Imaginary $0$-Nearest Neighbour and Its Improved Convergence Rate