作者: L. Devroye
关键词: Statistics 、 Nonparametric regression 、 Mathematics 、 Uniform convergence 、 Applied mathematics 、 k-nearest neighbors algorithm 、 Function (mathematics) 、 Estimator 、 Noise 、 Search algorithm 、 Simple random sample
摘要: A class of nonparametric regression function estimates generalizing the nearest neighbor estimate Cover [ 12] is presented. Under various noise conditions, it shown that are strongly uniformly consistent. The uniform convergence can be exploited to design a simple random search algorithm for global minimization function.