Scholarlink

<p>With the continued advancement of educational assessment systems, there is a growing demand for models capable of assessing student comprehensive quality in a holistic and objective manner. To accurately capture students’ multidimensional abilities and developmental potential, this study proposes an assessment system that integrates the Improved Analytic Hierarchy Process (IAHP) with a Back Propagation Neural Network (BPNN), referred to IAHP–BPNN. IAHP is employed to screen evaluation indicators and calculate their weights, addressing the issue of inconsistency in judgment matrices inherent in the traditional Analytic Hierarchy Process (AHP). The BPNN then trains on and predicts student data, enabling intelligent evaluation of student comprehensive quality. Experimental results demonstrate that the proposed model achieves an accuracy exceeding 90% on the test set, approximately 15% higher than the traditional AHP-based approach. Furthermore, on small-sample datasets, the hybrid model maintains strong performance, with a Mean Absolute Error of only 0.05, compared to over 0.15 using single-method models. These findings indicate that the IAHP–BPNN model effectively identifies strengths and weaknesses in student development and offers valuable guidance for educators.</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

A student comprehensive quality assessment system based on improved analytic hierarchy process and back propagation neural network

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年8月29日

查看全文
  • Lu Zhang

摘要

With the continued advancement of educational assessment systems, there is a growing demand for models capable of assessing student comprehensive quality in a holistic and objective manner. To accurately capture students’ multidimensional abilities and developmental potential, this study proposes an assessment system that integrates the Improved Analytic Hierarchy Process (IAHP) with a Back Propagation Neural Network (BPNN), referred to IAHP–BPNN. IAHP is employed to screen evaluation indicators and calculate their weights, addressing the issue of inconsistency in judgment matrices inherent in the traditional Analytic Hierarchy Process (AHP). The BPNN then trains on and predicts student data, enabling intelligent evaluation of student comprehensive quality. Experimental results demonstrate that the proposed model achieves an accuracy exceeding 90% on the test set, approximately 15% higher than the traditional AHP-based approach. Furthermore, on small-sample datasets, the hybrid model maintains strong performance, with a Mean Absolute Error of only 0.05, compared to over 0.15 using single-method models. These findings indicate that the IAHP–BPNN model effectively identifies strengths and weaknesses in student development and offers valuable guidance for educators.