Gold Price Prediction Using Machine Learning

Abstract

Gold price prediction is very essential for investment in gold market. Gold is an important economical factor for a developing country like Bangladesh. It is a precious material and Bangladesh is a traditional country where everyone uses gold mainly for jewelry. So, the demand of gold is endless. Our people use to invest in gold because its price is continuously changing. So that it takes a strong position in our national economy. There are so many factors that influence gold price such as Global Gold Market, United States Currency Rate and Geopolitical Risks. We are facing a pandemic of Covid-19 which affected the gold market worldwide. US currency have been a major factor for gold rate movement for many years. In this research we have been used daily USD value in BDT and Silver price from January 2015 to September 2021 as two main factors for predicting gold price. In this work we tried to implementsupervised machine learning algorithms such as Random Forest Regression, KNN, Decision Tree Classifier, Logistic Regression, and Multinomial NB to find out appropriate algorithm for our work. We have compared all these algorithms in our study. From this study, Random Forest Regression performed best among all other algorithms.