Applying multi-criteria decision classifier in multi-class classification
Chen‐Tung Arthur Chen, Ping‐Feng Pai, Wei-Zhan Hung · 2011
Multi-criteria decision-making (MCDM) is one of the most widely used decision methodologies in the sciences, business, government and engineering fields. The goal of this research is to extend multi-criteria decision-making method to multi-class classification problem. Multi-criteria decision classifier (MCDC) is a new classifier and is proposed by this research for dealing with multi class classification problem. The application of MCDC is that it can consider as a weak classifier for adaboost classifier although the accurate rate of MCDC is worst than back-propagation neural network (BPNN). This study will make an example to implement and compare the proposed method with BPNN, KNN and Bayesian classifier. Finally, we will make a conclusion and future research.