A Stock Trading System based on Moving Average Patterns and Turning Point Matrix

In this paper, an intelligent stock trading system utilizing moving average patterns which reflect short term price changes and turning point matrix which represents the long term history of stock price. Separated predictors are learned for the selected four patterns including the golden cross. In addition, the proposed system defines and utilizes turning point matrix which succinctly represents the long term history of stock price by Fibonacci numbers. Through experimental results on KOSPI, it is shown that the use of the pattern-based multiple predictors and turning point matrix can contribute to enhancement of the final trading performance.