Building an Associative Classifier Based on Fuzzy Association Rules

Guoqing Chen,Zuoliang Chen

doi:10.2991/ijcis.2008.1.3.7

Abstract

Classification based on association rules is considered to be effective and advantageous in many cases. However, there is a so-called sharp boundary problem in association rules mining with quantitative attribute domains. This paper aims at proposing an associative classification approach, namely Classification with Fuzzy Association Rules (CFAR), where fuzzy logic is used in partitioning the domains. In doing so, the notions of support and confidence are extended, along with the notion of compact set in dealing with rule redundancy and conflict. Furthermore, the corresponding mining algorithm is introduced and tested on benchmarking datasets. The experimental results revealed that CFAR generated better understandability in terms of fewer rules and smother boundaries than the traditional CBA approach while maintaining satisfactory accuracy.

Highlights

Classification and association rule mining[23] are two major areas of research and applications nowadays in knowledge discovery
This paper is aimed at dealing with the “sharp boundary” problem for quantitative domains and providing an approach to building an associative classification based on fuzzy association rules
The second part is to discuss the impact of threshold MS on Classification with Fuzzy Association Rules (CFAR) outcomes

Summary

Introduction

Classification and association rule mining[23] are two major areas of research and applications nowadays in knowledge discovery. An association rule (AR) is of the form, X ⇒ Y where X and Y are sets of data items. The goal of association rule mining is to generate certain associative relationships between data items with the degrees of confidence and support greater than user specified thresholds. A typical association rule application is market baskets analysis describing, for example, the customers’ buying behavior such as “Fruit ⇒ Meat” meaning that customers who bought fruit tended to buy meat, which reflects association between occurrences of data items. 3,6,8,12,14,17,18,21,29,32,33 A worth-noting type of approaches is classification based on association rules, aimed at building a classifier by discovering a small set of rules to form a so-called associative classifier. An associative classifier is composed of only those association rules of the form

Objectives

Results

Conclusion