A FUZZY-BASED ROUGH SETS CLASSIFIER FOR FORECASTING QUARTERLY PGR IN THE STOCK MARKET (PART II)

作者：Cheng Ching Hsue; Chen Yoush Yang^*

来源：International Journal of Innovative Computing Information and Control, 2011, 7(3): 1209-1228.

• Rough set theory
• Fuzzy-rule similarity
• Profit growth rate (PGR)
• Feature selection
• Discretization method
• Minimize entropy principal approach (MEPA)

Three hybrid models are proposed for solving the practical problems of forecasting quarterly PGR (profit growth rate) and based on six essential components: experiential knowledge (EK), feature selection method (FSM), discretization method (DM), fuzzy set theory (FST), rule filter (RP) and rough set theory (RST). Following the Part I (the previous paper), a called PGR dataset collected from the financial holding stocks in Taiwan's stock market is implemented as the empirical case study in the Part, II (this paper) to evaluate the proposed hybrid models. An external comparison and internal comparison are conducted; concurrently, some findings and management implications are disclosed from the experimental results. The results include decision rules of a set to directly rule the strategy of investment and intelligently offer a powerful explanation for the investors. They are of value to both academicians and practitioners.

• 出版日期2011-3