Socioeconomic and Accessibility Predictors of Girls' School Dropout: A Machine Learning Analysis

Shubhii Shuklla

Abstract: Girls dropping out of education is becoming a global problem, which is happening due to economic and environmental conditions. In this study, the researcher focused on parental educational background, distance travelled to commute to school daily, to analyse the dropout issue. The study has considered four classification algorithms: logistic regression, decision tree, random forest and support vector machine to predict the students' continuation in school using the UCI dataset of student performance. During the study, the author has identified that logistic regression has shown the most prominent result of 67% and support vector machine has shown 0.58 in terms of ROC and AUC. The study highlighted that girl students are more likely to drop out if their parents have a lower educational background, and the impact of decision-making based on data-driven insights has a direct impact on girl students? dropout reduction rate.

Keywords: Female Education, School Dropout, Machine Learning, Mother Education, Father Education, School Distance, Random Forest, Predictive Analytics

How to Cite?: Shubhii Shuklla, "Socioeconomic and Accessibility Predictors of Girls' School Dropout: A Machine Learning Analysis", Volume 14 Issue 12, December 2025, International Journal of Science and Research (IJSR), Pages: 2147-2151, https://www.ijsr.net/getabstract.php?paperid=SR251224165536, DOI: https://dx.doi.org/10.21275/SR251224165536