Overview and Analysis of Contrast Pattern Based Classifica- tion

This chapter presents a high-level overview and analysis of contrast pattern (CP) based classification. It identifies two main issues for CP-based classification, namely CP model selection and CP-based classification strategy. It describes and compares representative CP-based classification algorithms, with respect to how they deal with the two main issues. It also discusses how and why CP-based classifiers can achieve high classification accuracy. Together, the presented algorithms use many of the important techniques that have been introduced in the general CP-based classification paradigm.