Nearest Neighbor Classification Rule for Mixtures of Discrete and Continuous Random Variables

✍ Scribed by T. J. Wojciechowski

• Publisher: John Wiley and Sons
• Year: 1987
• Tongue: English
• Weight: 297 KB
• Volume: 29
• Category: Article
• ISSN: 0323-3847
• DOI: 10.1002/bimj.4710290808

✦ Synopsis

In this paper very simple nonparametric c l d i c a t i o n rule for mixtures of discrete and oonhuons random variables is described. It ie based on the method of neatest neighbor proposed by COVEB and HABT (1967). The bounds on the limit of the near& neighbor ruleriske are given. Both lower and upper bound depend on the Beyes risk and the lose function. Finally the method ia compared with other existing methods on Borne practical date set.