Fast and Accuracy Control Chart Pattern Recognition using a New cluster-k-Nearest Neighbor

By taking advantage of both k-NN which is highly accurate and K-means cluster which is able to reduce the time of classi¯cation, we can introduce Cluster-k-Nearest Neighbor as "variable k"-NN dealing with the centroid or mean point of all subclasses generated by clustering algorithm. In...

Full description

Saved in:
Bibliographic Details
Main Author: Brahim Belhaouari, samir
Format: Citation Index Journal
Published: 2008
Subjects:
Online Access:http://eprints.utp.edu.my/905/1/C_K_NN.pdf
http://eprints.utp.edu.my/905/
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:By taking advantage of both k-NN which is highly accurate and K-means cluster which is able to reduce the time of classi¯cation, we can introduce Cluster-k-Nearest Neighbor as "variable k"-NN dealing with the centroid or mean point of all subclasses generated by clustering algorithm. In general the algorithm of K-means cluster is not stable, in term of accuracy, for that reason we develop another algorithm for clustering our space which gives a higher accuracy than K-means cluster, less subclass number, stability and bounded time of classi¯cation with respect to the variable data size. We ¯nd between 96% and 99.7 % of accuracy in the classi¯cation of 6 di®erent types of Time series by using K-means cluster algorithm and we ¯nd 99.7% by using the new clustering algorithm.