Future Gold Price Prediction Using Ensemble Learning Techniques and Isolation Forest Algorithm

Abstract

Previously, gold is a mineral with unique characteristics that draw people's attention for a variety of reasons, including its high demand in jewelry. In addition to other forms of payment, gold was utilized to finance commercial transactions all over the world. Moreover, gold is a national financial asset, and different states preserved and grew their gold holdings, indicating that they were affluent and forward-thinking. Across the globe, central banks keep precious metals such as gold for loans and other requirements of the people. In India and other parts of the globe, gold can be used to offer as a compliment. Gold prices are heavily influenced by global commodity demand and supply. Inside this article, we are going to predict the gold rate for the next 30 days and also find the predictions of the financial year 2020–21, next fiscal year 2021–22, and until the middle of the following fiscal year 2022–23 using ensemble learning technique and isolation forest algorithm. We used information from the Internet to create predictions using the ensemble learning method and unsupervised-based approach. In this paper, we have forecasted the predictions of gold rate using different algorithms which will give better accuracy based upon above techniques. And also by applying these techniques in the whole gold dataset, we are going to compare the accuracy using ensemble and isolation forest algorithms which will give better results at all.