Search

Applying the association rule into classification can improve the accuracy and obtain some valuable rules and information that cannot be captured by other classification approaches. However, the rule generation procedure is very time-consuming when encountering large data set. Besides, traditional classifier building is organized in several separate phases which may also degrade the efficiency of these approaches. In this paper, a new class based associative classification approach (CACA) is proposed. The class label is taken good advantage of in the rule mining step so as to cut down the searching space. The proposed algorithm also synchronize the rule generation and classifier building phases, shrinking the rule mining space when building the classifier to help speed up the rule generation. Experimental result suggested that CACA is making better performances in accuracy and efficiency in Associative classification approaches. [ABSTRACT FROM AUTHOR]

The XCS classifier system has shown to solve typical classification problems competitively to other machine learning algorithms. This paper is to introduce a flexible classifier selection method in which mach-set generation is performed with respect to the system performance during the training. It also presents a complementary part for the Search Component of the classifier system in order to create new classifiers regarding the input message from the environment, and previous knowledge. The modified XCS learns more reliable and performs obviously better than traditional XCS. [ABSTRACT FROM AUTHOR]