ABSTRACT
This chapter introduces a multiple rules–based decision making (MRDM) model. It provides the details of hybrid bipolar model and a numerical example of improving the financial performance (FP) of semiconductor companies. This bipolar decision model has the advantages of MRDM in retrieving understandable knowledge to support decision making; in addition, it highlights the importance of contexts to performance evaluation and improvement planning. The bipolar approach can be categorized into the emerging field of MRDM, which attempts to provide insightful and transparent analyses for businesses to plan for performance improvements. The hybrid model comprises multiple soft computing techniques: dominance-based rough set approach, fuzzy evaluation, and the threshold-based bipolar model. These techniques bridge the imprecise judgments of human beings with induce rough knowledge from historical data, which is crucial to modeling practical business dynamics. Based on the proposed taxonomy of business analytics, it could be categorized into three types of analytics: descriptive, predictive, and prescriptive.
