Data Mining by Attribute Decomposition with Semiconductor Manufacturing Case Study

Abstract

This chapter examines the Attribute Decomposition Approach with simple Bayesian combination for dealing with classification problems in the semiconductor industry. Often classification problems in this industry contain high number of attributes (due to the complexity of the manufacturing process) and moderate numbers of records. According to the Attribute Decomposition Approach, the set of input attributes is automatically decomposed into several subsets. A classification model is built for each subset, then all the models are combined using simple Bayesian combination. This chapter presents theoretical and practical foundation for the Attribute Decomposition Approach. A greedy procedure, called D-IFN, is developed to decompose the input attributes set into subsets and build a classification model for each subset separately. The algorithm has been applied to problems in the semiconductor industry and on variety of databases from other application domains. The results achieved in the empirical comparison testing with well-known classification methods (like C4.5) indicate the superiority of the decomposition approach.