Comparison of machine learning methods for stock return series analysis

Jian Wang (FSU)

Time: 3:35 pm, Room: LOV 201

Abstract: Machine learning is a method of teaching computers to make predictions based on data. It is one important branch of artificial intelligence research area. Nowadays, those popular algorithms have been widely used in our financial engineering research area, for example, prediction of stock trend or limit order book prices. In this presentation, we propose the use of 452 daily stock return series historical data as input feature to some machine learning algorithms and conduct both variable selection and prediction jobs. The methodologies we used were Bayesian network, Logistic regression, Lasso, and Support vector machine under different kernel functions. Besides, we also compared the efficiency and accuracy rate of machine learning methods based on two different languages(Python and R) and showed that python had more advantages in this case.