Paper Details: Downloads: 575
Serial Number: P1121101448
Title: Classification Assessment based on Accuracy, Compactness, and Speed of C4.5 and CPAR: A Comparative Study
Authors: H. Rahmat and M.S.M. Said and N.A. Amit and N.Z.M. Yasin and A. Mustapha
Abstract: Classification is a widely used data mining task in organizing business operations. The classification task, used in conjunction with the association rule mining task form an associative classification. This paper investigates performance of the associative classifier CPAR against the traditional classifier C4.5. The experiments use UCI zoo dataset with different attribute dimensions in measuring the speed, accuracy, and compactness of two classifiers CPAR and C4.5. The experiments show that CPAR algorithms outperform C4.5 in all aspects measured and are at higher performance as compared to previous results.
Keywords: Classification, Association-based classification, Predictive Association Rules
Journal/Conference: ICGST Conference on Artificial Intelligence and Machine Learning, AIML-11
Volume:
Issue:
Submission Date: 1/3/2011 12:00:00 AM
Review Date: 3/27/2011 12:00:00 AM
Publishing Date: 4/6/2011 12:00:00 AM
Article Downloads: 575
Download:

Facebook