Prediction of Stock Trading Signal Based on Support Vector Machine

Abstract

The prediction of stock trading signal is studied in this paper. Considering the excellent performance of Support Vector Machine (SVM) in pattern recognition, we apply SVM to construct a prediction model to find the stock trading signal. In addition, Piecewise linear representation (PLR) is good at extracting valuable information from a time sequence. PLR is used for checking of turning points in this study. The experiments on some real stocks show that SVM obtains a better result in prediction accuracy and profitability than traditional Back Propagation neural network does.