Associative classification, which uses association rules for classification, has achieved high accuracy in comparison with other classification approaches. However, the confidence measure, which is used to select association rules for classification, may not conform to the prediction accuracy of the rules. In this paper, we propose a measure for association rules called prediction confidence to measure the accuracy of the prediction of association rules. An approach for estimating the prediction confidence of a rule is also given. The use of prediction confidence instead of confidence measure helps gather better association rules for associative classification. As a result, a more accurate associative classifier can be constructed using the prediction confidence measure.
[1]
Wynne Hsu,et al.
Integrating Classification and Association Rule Mining
,
1998,
KDD.
[2]
Nils J. Nilsson,et al.
MLC++, A Machine Learning Library in C++.
,
1995
.
[3]
Ramakrishnan Srikant,et al.
Fast algorithms for mining association rules
,
1998,
VLDB 1998.
[4]
Bing Liu,et al.
Classification Using Association Rules: Weaknesses and Enhancements
,
2001
.
[5]
Jian Pei,et al.
CMAR: accurate and efficient classification based on multiple class-association rules
,
2001,
Proceedings 2001 IEEE International Conference on Data Mining.