Historically and globally, there has been a notable underrepresentation of women in STEM (science, technology, engineering, and mathematics) careers. This disparity is influenced by various factors, including gender stereotypes and women’s self-perception of their academic abilities. The present systematic review aims to explore the application of machine learning techniques to predict women’s entry into STEM careers, with a focus on identifying and analyzing socioeconomic and environmental factors that influence their decision-making. The review was conducted using academic repositories such as Scopus and Web of Science, resulting in the selection of 40 articles. These studies included educational records, surveys conducted with students, preprocessing techniques, machine learning algorithms, and metrics for information analysis. Key findings revealed that the support vector machine (SVM) model was the most commonly used, while random forest (RF) and gradient boosting (GB) algorithms demonstrated superior performance. Regarding validation metrics, the majority of studies utilized accuracy, precision, F1 score, recall, and ROC curve to assess model effectiveness. In conclusion, the present review provides a comprehensive understanding of the gender gap in STEM fields, highlighting the significant influence of specific factors and offering insights into predictive modeling for women’s participation in these areas.