Price Predictions Based on Evolvable Strategies, Principal Component Analysis, and Randomness of Data Strings
Mieko TANAKA-YAMAWAKI*, Xin YANG*, and Ryota ITOI*
*Department of Information and Electronics, Graduate School of Engineering, Tottori University, 4-101 Koyamacho-Minami, Tottori
This paper summarizes the trials of predicting price movements in the financial market in various ways. The first example is to use the framework of the game theory between two players by regarding the next price movement as the next action of a player and the information of the environment surrounding the price movement as the action of the opponent player by using the evolutional computation. The second example is to predict trendy sectors by using RMT-PCA. The third example is to use the randomness of the price fluctuations by means of the RMT-test. An empirical rule “High randomness predicts a good performance in the next period” is extracted by studying the randomness of stocks in the period of 2007-2009.