Forecasting of Gold Prices Using Bayesian Regularization Neural Network

Abstract

Prediction of gold price has been done by many, the investors, artificial intelligence experts, machine learning experts and stock market agents, etc. Ten years gold price from January 2008 to May 2018 has been taken for prediction in this study using Bayesian neural network. Bayesian neural network is selected for prediction of price of gold because it is found suitable for time series data and does not depend on any historical feature. The forecasted price is also compared with actual price to find percentage error in predicted values. Finally, mean percentage error is calculated, which is found 1% here.