Abstract:
The main aim of this paper is focused on the possibility of using SVM regression model as a decision making tool. This model is used for financial markets prediction (generating buy and sell trading orders), especially for U.S. stocks and ETFs (AAPL, IBM, SPY, GLD, SLV, OIL, XIV). This portfolio of U.S. stocks and ETFs is used for scientific research and for testing in real market, too. The main assumptions were to predict Friday’s closing prices on Monday (before market opens). We express the excitement about this statistical technique as it performed admirably even without optimising the input variable set or the regression parameters. We used the conventional RMSE as a measure of this uncertainty, but we also noted slight ambiguity of its interpretation.
