Scholarlink

<p>The rapid growth of educational technologies has generated extensive data on student performance. This study introduces an enhanced grade prediction method that groups students using spectral clustering optimized with a KD-tree. By leveraging the structural properties of educational data, our approach forms more coherent student clusters, leading to richer input features for prediction models. Evaluated on diverse datasets, the method demonstrates significant improvements over conventional techniques. The results highlight its potential to support more accurate grade prediction and contribute to personalized education.</p>

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

Clustering students with spectral algorithm enhanced by KDtree to improve test score predictions

  • 期刊论文

  • 发表日期：

    2025年10月21日

查看全文
  • Abderrafik Laakel Hemdanou,
  • Youssef Achtoun,
  • Ismail Tahiri,
  • Mohammed Lamarti Sefian

摘要

The rapid growth of educational technologies has generated extensive data on student performance. This study introduces an enhanced grade prediction method that groups students using spectral clustering optimized with a KD-tree. By leveraging the structural properties of educational data, our approach forms more coherent student clusters, leading to richer input features for prediction models. Evaluated on diverse datasets, the method demonstrates significant improvements over conventional techniques. The results highlight its potential to support more accurate grade prediction and contribute to personalized education.