Foreign Currency Exchange Rate Prediction Using Long Short-Term Memory, Support Vector Regression and Random Forest Regression

Abstract

This chapter aims to predict the foreign currency exchange rate on the basis of the US dollar over twenty-two different currencies. This chapter proposes two machine learning algorithms like Support Vector Regression (SVR), Random Forest Regression (RFR) and one deep learning algorithm named Long Short-Term Memory (LSTM), for the technical analysis of currency exchange rate prediction. The authors use Mean absolute error (MAE), Mean squared error (MSE), Root Mean Squared Error (RMSE), and mean absolute percentage error (MAPE) to measure the performance of the algorithms. Empirical findings specify that the overall performance of the algorithms is outstanding, but the Long Short-Term Memory (LSTM) shows less error than others. This study is useful for the stakeholders to set a wide range of approaches for the foreign exchange market.KeywordsCurrency exchange rate predictionMachine learningDeep learningTime series analysis