Gold Price Prediction using Ensemble based Machine Learning Techniques

Abstract

This article is based on a study conducted to understand the relationship between gold price and selected factors influencing it, namely stock market, crude oil price, rupee dollar exchange rate, inflation and interest rate. Monthly price data for the period January 2000 to December 2018 was used for the study. The data was further split into two periods, period I from January 2000 to October 2011 during which the gold price exhibits a raising trend and period II from November 2011 to December 2018 where the gold price is showing a horizontal trend. Three machine learning algorithms, linear regression, random forest regression and gradient boosting regression were used in analyzing these data. It is found that the correlation between the variables is strong during the period I and weak during period II. While these models show good fit with data during period I, the fitness is not good during the period II. While random forest regression is found to have better prediction accuracy for the entire period, gradient boosting regression is found to give better accuracy for the two periods taken separately.