Gender disparity within Science, Technology, Engineering, and Mathematics (STEM) programs remains a global challenge, affecting women’s academic and professional opportunities. This systematic review employs the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines to evaluate machine learning (ML) techniques aimed at predicting gender disparities in STEM courses. A systematic literature review process was employed to map findings from studies published between 2016 and 2024, utilizing major scientific databases such as Science Direct, Scopus, IEEE Xplore, ACM, and Springer. Following rigorous selection procedures, 58 papers were chosen for analysis, revealing that only 49 papers specifically addressed ML approaches. The study examines the strengths and weaknesses of various ML techniques, aiming to determine the most appropriate ML prediction model for this research area. Key findings and trends in the use of ML for addressing gender inequality are highlighted, offering insights for evidence-based interventions and policies in STEM education and professional fields. The synthesis of existing literature provides a roadmap for future research. Leveraging advanced algorithms and comprehensive data analysis, we can gain deeper insights into the factors driving these disparities and work towards creating a more inclusive and equitable STEM environment.