The Application of Random Forest in Finance

Abstract

The variation of futures price is affected by a lot of factors. It is a challenge to predict the price’s trend. In this paper, we apply random forest technique to predict the type of K-line in next day. First, we empirically select nine technical indexes as input variables or features. Then, these features and the associated class label, i.e. the type of K-line, are used to construct RF classifier. The experimental results demonstrate that RF is effective and can be used in the trend prediction of futures price.