Stock movement prediction using fuzzy Intertransaction Class Association Rule Mining based on Genetic Network Programming

Intertransaction class association rule mining (CARM) is an efficient method to predict the stock movement using the data of many stocks within a few days. And a crisp intertransaction CARM method based on genetic network programming (GNP) has been studied in our previous study. In this paper, a fuzzy intertransacion CARM method is presented to reduce the loss of information in discretization and obtain more profitability with less risk in the stock investment. The proposed method consists of two steps: fuzzy intertransaction rule extraction and classifier building. We applied the proposed method to Tokyo stock exchange market and compared its experimental results with the crisp case as well as some other methods.