… Extracting Trading Rules from the Multiple Classifiers and Technical Indicators in Stock Market Kyoung … than other comparable models and buy-and-hold strategy … 3.1 Domain Knowledge based …

Abstract. This study intends to mine reasonable trading rules by classifying the up/down fluctuant direction of the price for Korea Stock Price Index 200 (KOSPI 200) futures. This research consists of two stages. The first stage is classifying the fluctuant direction of the price for KOSPI 200 futures with several technical indicators using artificial intelligence techniques. And the second stage is mining the trading rules to resolute conflict among the outputs of the first stage using the inductive learning. To verify the effectiveness of proposed approach, this study composes four comparable models and performs statistical test. Experimental results show that the classification performance of the proposed model outperforms that of other comparable models. In addition, the proposed model yields higher profit than other comparable models and buy-and-hold strategy. Key words: Rule extraction, Conflict resolution, Inductive learning, Artificial neural network, Case-based reasoning. Introduction. It is known that stock market shows a nonlinear pattern. By this reason, there may be some limitations to analyze the stock market with linear model. Therefore, artificial intelligence (AI) techniques are often used for the nonlinear pattern analysis such as stock market analysis. Several studies show that artificial neural network and case-based reasoning can be applied to stock market predictions, but most of them have focused mainly on prediction of spot market and time series data (Ahmadi, 1990, Kamijo and Tanikawa, 1990, Kimoto et al., 1990, Yoon and Swales, 1991). They also did not bare outstanding prediction accuracy because tremendous noise and non-stationarity of data. Korea launched trading in index futures market (KOSPI 200) on May 1996, then more people became attracted to this market, because the investor can avoid or reduce the risk of stock investment via index futures. The lack of the research on the index futures market in Korea has not met the people’s interest. By this reason, this research intends to classify the daily up/down fluctuant direction of the futures price for Korea stock index (KOSPI 200) to meet this recent surge of interest…

Source: http://afis.kaist.ac.kr/

Download Extracting Trading Rules from the Multiple Classifiers and Technical Indicators in Stock Market