Theoretical And Empirical Criteria For The Edited Nearest Neighbour Classifier
2015 IEEE International Conference on Data Mining(2015)
摘要
We aim to dispel the blind faith in theoretical criteria for optimisation of the edited nearest neighbour classifier and its version called the Voronoi classifier. Three criteria from past and recent literature are considered: two bounds using Vapnik-Chervonenkis (VC) dimension and a probabilistic criterion derived by a Bayesian approach. We demonstrate the shortcomings of these criteria for selecting the best reference set, and summarise alternative empirical criteria found in the literature.
更多查看译文
关键词
edited nearest neighbour classifier,theoretical criteria,empirical criteria,Voronoi classifier,Vapnik-Chervonenkis dimension,probabilistic criterion,Bayesian approach
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络