A Comparative Study of the Application of Machine Learning Techniques to Analysis and Prediction of the Market Prices for Precious Group Metals

For efficient management of operations, mining companies are required to be able to predict potential scenarios for main market factors. Competitive advantage and value creation are established through managing the value chain well. The mining value chain is usually faced with challenges from various shifts in commodity markets as well as ongoing uncertainties. Therefore, there is a need for accurate analysis of data from the markets, for the prediction of commodities, for instance, Precious Group Metals (PGM). The main objective of this research work is to apply machine learning methodologies with the aim to identify patterns and trends in the Precious Group Metals, and thereafter analyze and predict the future trends of these markets. Numerous machine learning models are applied to a dataset consisting of 15 years of daily price for eight precious metals namely Platinum, Palladium, Rhodium, Ruthenium, Osmium, Indium, Silver, and Gold. The results reveal the patterns and trends within the data sets and that different machine learning methods could be suited for the prediction of market prices for each of the PGMs. The application of an amalgamation of various machine learning methods may result in more accurate predictions of the PGM markets. Accurate prediction of market prices for PGMs will possibly result in unlocking untapped sources of value in the mining industry.