Performance Comparison Between Support Vector Regression and Long Short - Term Memory for Prediction of Stock Market

Stock prediction is one of the exciting solutions in the world of computing. A stock prediction has a vital role in determining the pattern of stock movements over time. A successful price prediction model is profitable for investors to determine the time to buy or sell stocks. In this study, we used Support Vector Regression and Long Short-Term Memory to predict stock price in Nasdaq's dataset. And then, we compare the performance result of both methods to show which method is the most accurate in producing the solutions. The results show that the LSTM model has a better result than the SVR model. The best LSTM model with RMSE value is 192.6, MAE of 144.12, and R-Squared of 0.9941. These results show good performance because they give the smallest error value and the largest R-Squared value.