Scholarlink

This research explores the application of machine learning in educational management for universities, specifically focusing on predicting student graduation or dropout rates. In the context of university management, predicting student outcomes such as graduation or dropout is crucial for recognizing students who are in danger early and implementing targeted action to support their educational journey. This research utilizes a comprehensive dataset sourced from Kaggle, encompassing various student attributes such as academic performance, demographic information, and socioeconomic factors. Through rigorous feature analysis and data preprocessing, this study lays the groundwork for applying classification algorithms, including logistic regression, AdaBoost, Gradient Boosting, and decision trees, to the dataset to improve predictive accuracy, Ensemble Methods such as Boosting, Bagging, and Voting classifiers are applied. Results reveal promising accuracy rates: Bagging and Boosting classifiers achieve 90% accuracy each, while the Voting classifier reaches 92%.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Enhancing Student Success: Predictive Modeling of Graduation and Dropout Rates in University Management Using Machine Learning

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Ashok Sapkota,
  • Sandeep Kaur

摘要

This research explores the application of machine learning in educational management for universities, specifically focusing on predicting student graduation or dropout rates. In the context of university management, predicting student outcomes such as graduation or dropout is crucial for recognizing students who are in danger early and implementing targeted action to support their educational journey. This research utilizes a comprehensive dataset sourced from Kaggle, encompassing various student attributes such as academic performance, demographic information, and socioeconomic factors. Through rigorous feature analysis and data preprocessing, this study lays the groundwork for applying classification algorithms, including logistic regression, AdaBoost, Gradient Boosting, and decision trees, to the dataset to improve predictive accuracy, Ensemble Methods such as Boosting, Bagging, and Voting classifiers are applied. Results reveal promising accuracy rates: Bagging and Boosting classifiers achieve 90% accuracy each, while the Voting classifier reaches 92%.