Gold Price Forecasting Using Regression Techniques for Settling Economic and Stock Market Inconsistency

Abstract

Gold is one of the most important metals in a country's economy. For financial banks and stock exchanges, it is a regulatory factor. Gold has the potential to have a huge impact on the economy. Gold price volatility is a common occurrence in almost all countries. Our own country, Bangladesh, is no different. Because it is held as a reserve by the central bank, changes in its price can cause problems in the country's economy. In this paper, we present our models for predicting gold prices on a daily basis. We have used machine learning approach for achieving this goal. To forecast the daily gold price, we used Support Vector Regression (SVR), Random Forest Regressor (RFR), Decision Tree, Gradient Boosting, and XGBoost models. All the models that we have generated produce outcomes that are much acceptable. Amidst all the models we devised Random Forest Regressor (RFR) has generated the best outcome in all phases. The accuracy attained by the RFR algorithm is around 99% in all cases