STOCK MARKET TREND PREDICTION USING SUPPORT VECTOR MACHINES

The aim of the paper was to outline a trend prediction model for the BELEX15 stock market index of the Belgrade stock exchange based on Support Vector Machines (SVMs). The feature selection was carried out through the analysis of technical and macroeconomics indicators. In addition, the SVM method was compared with a "similar" one, the least squares support vector machines - LS-SVMs to analyze their classification precisions and complexity. The test results indicate that the SVMs outperform benchmarking models and are suitable for short-term stock market trend predictions