A novel Chi2 algorithm for discretization of continuous attributes

Authors:
Wenyu Qu;Deqian Yan;Yu Sang;Hongxia Liang;Masaru Kitsuregawa;Keqiu Li
Affiliations:
School of Computer Science and Technology, Dalian Maritime University, Dalian, China and Institute of Industrial Science, The University of Tokyo, Tokyo, Japan;Department of Computer Science and Engineering, Liaoning Normal University, Dalian, China;Department of Computer Science and Engineering, Liaoning Normal University, Dalian, China;Department of Computer Science and Engineering, Liaoning Normal University, Dalian, China;Institute of Industrial Science, The University of Tokyo, Tokyo, Japan;Department of Computer Science and Engineering, Dalian University of Technology, Dalian, China
Venue:
APWeb'08 Proceedings of the 10th Asia-Pacific web conference on Progress in WWW research and development
Year:
2008

Citing 7
Cited 1

C4.5: programs for machine learning

C4.5: programs for machine learning
Feature Selection via Discretization

IEEE Transactions on Knowledge and Data Engineering
A Modified Chi2 Algorithm for Discretization

IEEE Transactions on Knowledge and Data Engineering
Class-Dependent Discretization for Inductive Learning from Continuous and Mixed-Mode Data

IEEE Transactions on Pattern Analysis and Machine Intelligence
CAIM Discretization Algorithm

IEEE Transactions on Knowledge and Data Engineering
An Extended Chi2 Algorithm for Discretization of Real Value Attributes

IEEE Transactions on Knowledge and Data Engineering
A Discretization Algorithm Based on a Heterogeneity Criterion

IEEE Transactions on Knowledge and Data Engineering

Compact classification of optimized Boolean reasoning with Particle Swarm Optimization

Intelligent Data Analysis

Quantified Score

Hi-index	0.00

Visualization

Abstract

The Chi2 algorithm, together with the Modified Chi2 algorithm and the Extended Chi2 algorithm, is famous for discretization algorithms with the base of probability and statistics. After studying these algorithms and analyzing their drawbacks, we present a new Chi2 algorithm called Rectified Chi2 algorithm, which regards a new merging standard as the basis of interval merging and discretizes the real value attributes exactly and reasonably. We also present a new sequence method (DSM) to overcome of the drawbacks that the Modified Chi2 algorithm and the Extended Chi2, i.e., they only adopt the maximal difference as standard of interval merger. We evaluate the performance of the Rectified Chi2 algorithm and DSM over extensive experiments. The experiment results show the effectiveness of the proposed algorithms.