Gold Prediction Based on XGBoost and OLS

Abstract

As a matter of fact, the value of gold is not only reflected in how much it is worth, but its unique nature also makes it determine economic and political policies. In reality, Gold is often seen as a hedge against inflation, influencing the direction of countries and investors. Gold prices from 2012-2018 from various websites were chosen as the dataset, including opening and closing prices and so on. With this in mind, using XGBoost and linear regression models successfully predicted the price of gold more accurately and it is not much different from the actual price. However, these methods are outdated, and geopolitical factors, national economic policies, and so on are some of the major factors affecting the price of gold in an era of increasingly up-to-date information. According to the analysis, the usage of integrated forecasting models combining Random Forests, MLR, SVM and ANN is a more futuristic way to predict the price of gold.