Analyzing Factors Affecting Female Student’s Success in Computer Science and Engineering

Abstract

Women have made phenomenal accomplishments in Computer Science and Engineering (CSE), and their interests have brought much-needed benchmarks to eradicate both gender differences and priorities in diversity in education and STEM fields around the tech sector they might be interested in. But all academic, social, and whatever might be the power of a potential woman is so small. Cultural obstacles in this area, costing families, societies and even governments in conditions of creative and economic development in a dynamic world. The present research efforts to determine and examine the psychosocial, contextual, and early educational influences which encourage the academic achievement of female CSE students. It considers the support of the parents, such, learning environment, cultural perceptions, self-discipline, self-confidence and so on. motivational variables that form part of improved academic performance. The research attempts to determine the accretion of these factors in the measure of the achievements or efforts encountered. by women. Data were collected on this using mixed methods. Primary data came from designed survey used on purpose on CSE students (women) of the Bangladeshi University, that provided a productive qualitative and quantitative experience information. Machine learning systems. Voting Classifier, Gradient Boosting, Random Forest, Stacking, were used to make analysis. XGBoost, Classifier, Logistic Regression, and SVC. Their work was assessed in terms of not only predicting the academic performance of the female students but also analyzing which factors. affected in no small measure on that success or failure. Give way to Motivation, self-discipline, a supportive One of the best predictors was the learning environment, and positive family reinforcement. of an academic success. Indeed, the aspects that seemed most related to poor performance in students were low self-esteem, poor command of English, and the pressure of part-time jobs, among others, always, gender stereotype. Among machine learning models used, Gradient Boosting emerged with the. Random Forest and Logistic Regression models having the highest prediction accuracy at 85%. On the other hand, the forecasting ability of Gradient Boosting and XGBoost here was low. This research, with the use of machine learning and a deep socio-cultural insight, lays a firm foundation. groundwork towards addressing the gender gap in STEM learning. Elicit responses that will increase teachers, schools, and policymaker’s awareness in different fields to special actions that would permit increasing early retention to such a level that gender equity and academic resilience of female students eventually achieve. Holistically, the research advocates for the nourishment of human-centered approaches with advanced technologies to ensure meaningful gains for female CSE students in realizing their full potential as they contribute towards the future of technology and innovation.