New prototype selection rule integrated condensing with editing process for the nearest neighbor rules

A new prototype selection method was proposed to reduce the training set for the nearest neighbor rule. The method aimed to integrate the advantage of editing and condensing method, and it only selects the points in class boundary into prototype set. It used information contained in internal points...

Full description

Saved in:
Bibliographic Details
Published in2005 IEEE International Conference on Industrial Technology pp. 950 - 954
Main Authors Yuangui Li, Jinjie Huang, Weidong Zhang, Xiaolei Zhang
Format Conference Proceeding
LanguageEnglish
Published IEEE 2005
Subjects
Automation
Density functional theory
Error analysis
Genetic algorithms
High performance computing
Nearest neighbor searches
Neural networks
Prototypes
Online AccessGet full text

Cover

More Information
Summary:A new prototype selection method was proposed to reduce the training set for the nearest neighbor rule. The method aimed to integrate the advantage of editing and condensing method, and it only selects the points in class boundary into prototype set. It used information contained in internal points to `clean' or edit overlapping and noise of training set, then condensing process which only keeps border points was used to obtain prototype set. Computational results show that it can obtain satisfactory performance and higher condensing rate than popular instance reduction algorithm
ISBN:0780394844
9780780394841
DOI:10.1109/ICIT.2005.1600772