This study investigates the challenge of predicting daily palladium prices in the United States using time series data from January 5, 1977, to March 26, 2024. The research employs Gaussian process regression algorithms, optimized through cross-validation and Bayesian optimization, to generate price predictions. The model achieves a relative root mean square error of 0.4598%, demonstrating reasonable accuracy in out-of-sample predictions from March 24, 2017, to March 26, 2024. The results suggest that price prediction models can provide valuable information for informed decision-making in the palladium industry. Commodity price forecasting is crucial for governments and investors to make well-informed strategies and reduce uncertainty. Previous research has focused on complex models like vector autoregressive (VAR), autoregressive integrated moving average (ARIMA), and vector error correction (VECM) models. However, recent advancements in artificial intelligence and machine learning have shown great potential for predicting commodity price trends. This study contributes to the growing body of research exploring machine learning techniques for commodity price forecasting, including applications in agriculture and energy sectors. The use of Gaussian process regression, combined with Bayesian optimization and cross-validation, offers a promising approach for accurate and reliable price predictions. The findings highlight the importance of time series forecasting in managing risks and making strategic decisions in the commodity market.This study investigates the challenge of predicting daily palladium prices in the United States using time series data from January 5, 1977, to March 26, 2024. The research employs Gaussian process regression algorithms, optimized through cross-validation and Bayesian optimization, to generate price predictions. The model achieves a relative root mean square error of 0.4598%, demonstrating reasonable accuracy in out-of-sample predictions from March 24, 2017, to March 26, 2024. The results suggest that price prediction models can provide valuable information for informed decision-making in the palladium industry. Commodity price forecasting is crucial for governments and investors to make well-informed strategies and reduce uncertainty. Previous research has focused on complex models like vector autoregressive (VAR), autoregressive integrated moving average (ARIMA), and vector error correction (VECM) models. However, recent advancements in artificial intelligence and machine learning have shown great potential for predicting commodity price trends. This study contributes to the growing body of research exploring machine learning techniques for commodity price forecasting, including applications in agriculture and energy sectors. The use of Gaussian process regression, combined with Bayesian optimization and cross-validation, offers a promising approach for accurate and reliable price predictions. The findings highlight the importance of time series forecasting in managing risks and making strategic decisions in the commodity market.