Palladium Price Predictions via Machine Learning

Predictions of prices for a wide variety of commodities have been relied upon by governments and investors over the course of history. The purpose of this study is to investigate the difficult challenge of predicting daily palladium prices for the United States by utilizing time series data ranging from January 5, 1977, to March 26, 2024. When it comes to this crucial evaluation of commodity prices, estimates have not been given sufficient consideration in earlier research. In this context, price predictions are generated by the utilization of Gaussian process regression algorithms, which are estimated through the utilization of cross-validation processes and Bayesian optimization approaches. With a relative root mean square error of 0.4598%, our empirical prediction approach produces price estimates that are generally accurate for the out-of-sample phase that spans from March 24, 2017, to March 26, 2024. In order to make educated choices about the palladium industry, governments and investors can utilize price prediction models to get the information they need.