A Statistical Feature Data Mining Framework for Students' Progress over Gender Variations

Abstract

Gender differences in academic progression are examined in this article using two novel indicators. Two indicators are used to show how long students have studied above the minimal amount of time needed to graduate. We examine these characteristics using Kaplan–Meier techniques, Expedited Fault Timeframe, Regression trees, Multiple Regression analysis tree branches, and a redesigned Gender Parity Index. It is possible to analyse data that has been censored as well as add other factors into the study using this unique mix of statistical methodologies. Gender disparities in university education may be studied using a combination of survival analysis and the Accelerated Failure Time (AFT) model, according to data from a Greek as well as Italian institution. It is possible to quantify the relevance of these influencing factors using data mining approaches such as survival trees and multivariate regression trees. The proposed indicators concurrently use the Multivariate Regression Tree technique, which considers more than one consistent outcome variable. Gender plays an essential part in this research, as women outperform males in terms of these new metrics by completing their studies in a shorter amount of time and with a greater level of performance, regardless of other student characteristics. The findings were improved across the board when using a gender parity index that was tweaked.