The instability of Premium Motor Spirit (PMS/Petrol) and other energy commodities prices, occasioned by volatile and dynamic movement of prices has been found to affect the cost of production in Nigeria. As a result, this paper studies current literature on the applications of machine learning techniques in forecasting PMS (Petrol) and other energy commodities prices. The review has been done through an electronic search of the published papers in the last 4 years (2019-2022). A total number of twenty-nine (29) publications on PMS (Petrol) and other energy commodities prices forecasting using machine learning models were selected for the study to identify research gaps and future works. This paper covers a summary of reviewed published papers on forecasting PMS (Petrol) and other energy commodities prices using machine learning techniques, semantic analysis of algorithms used, and the taxonomy of the models adopted in the published papers. The results showed that there are studies that presented the application of machine learning models in forecasting the prices of PMS (Petrol) and other energy commodities in other countries. However, very few of them have proposed the construction of machine learning models for forecasting PMS (Petrol) and other energy commodities prices in Nigeria. This leads to the need to develop new models, especially deep learning hybrid models.