This review explores the role of machine learning (ML) in predicting and optimizing product yields from plastic pyrolysis. It begins by outlining the fundamentals of plastic pyrolysis, including the behavior of various plastic types and key process variables such as temperature, catalysts, and residence time. The focus is on regression-based supervised learning models, including Support Vector Regression, Gaussian Process Regression, and ensemble tree methods. Applications of these models in enhancing oil, gas, and char yield predictions are critically discussed. The review concludes with insights into current challenges, data limitations, and recommendations for future research in ML-guided pyrolysis.