Predication of Futures Market by Using Boosting Algorithm

Abstract

AdaBoost is a machine learning algorithm based on boosting ideas. AdaBoost is the abbreviation of adaptive boosting, which is an algorithm for weak classifier to assemble as a strong classifier algorithm. There is a lot of data noise in finance market. In order to identify underlying trends in futures market, we propose to use standardized technical indicators to forecast rise or fall and assemble these indicators by AdaBoost algorithm innovatively. We use AdaBoost algorithm to optimize the weight of technical indicators and a good predication result is received. This research shows that these weak classifier can filter data noise in futures market effectively and exhibits that machine learning can get a better analysis result based on the conventional finance engineering analysis methods. This research is meaningful for individual investors.