Forecasting Gold Prices in India using Time series and Deep Learning Algorithms
P. Sai Shankar1, M. Krishna Reddy2

1P Sai Shankar*, Department of Statistics, University College of Science (OU), Hyderabad (Telangana), India.
2Dr. M. Krishna Reddy, Department of Statistic, CVR College of Engineering,and Technology, Hyderabad (Telangana), India.

Manuscript received on April 24, 2021. | Revised Manuscript received on May 03, 2021. | Manuscript published on June 30, 2021.. | PP: 21-27 | Volume-10 Issue-5, June 2021. | Retrieval Number:  100.1/ijeat.D25370410421 | DOI: 10.35940/ijeat.D2537.0610521
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: The primary object of this paper is to compare the traditional time series models with deep learning algorithm.The ARIMA model is developed to forecast Indian Gold prices using daily data for the period 2016 to 2020 obtained from World Gold Council.We fitted the ARIMA (2,1,2) model which exhibited the least AIC values. In the meanwhile, MLP, CNN and LSTM models are also examined to forecast the gold prices in India. Mean absolute error, mean absolute percentage error and root mean squared errors used to evaluate the forecasting performance of the models. Hence, LSTM model superior than that of the other three models for forecasting the gold prices in India. 
Keywords: Gold Prices, Box-Jenkins Methodology, ARIMA,Lag Variables, MLP, CNN and LSTMModels