Article: The prediction model of higher vocational students' classroom participation based on the fusion of deep learning and support vector machine Journal: International Journal of Information and Communication Technology (IJICT) 2026 Vol.27 No.9 pp.1 - 17 Abstract: Student engagement in vocational classrooms is a critical metric for assessing teaching effectiveness and talent development. To address the limitations of conventional assessment methods, we propose a hybrid deep learning-support vector machine (SVM) model for predicting participation levels. The approach integrates convolutional neural networks (CNN) and long short-term memory (LSTM) networks to extract high-dimensional temporal features from classroom videos and behavioural logs. These features are combined with traditional statistical indicators and classified using SVM through a feature-level fusion strategy. Evaluated on simulated vocational classroom data, the fused model achieves 92.3% accuracy and an F1-score of 0.914, significantly outperforming standalone CNN-LSTM or SVM models. This model enables real-time, quantitative assessment of classroom engagement and supports timely teaching interventions. Inderscience Publishers - linking academia, business and industry through research

Title: The prediction model of higher vocational students' classroom participation based on the fusion of deep learning and support vector machine

Authors: Yixuan Qiang

Addresses: School of Marxism, Jiangsu Vocational Institute of Commerce, Nanjing, 211168, China

Abstract: Student engagement in vocational classrooms is a critical metric for assessing teaching effectiveness and talent development. To address the limitations of conventional assessment methods, we propose a hybrid deep learning-support vector machine (SVM) model for predicting participation levels. The approach integrates convolutional neural networks (CNN) and long short-term memory (LSTM) networks to extract high-dimensional temporal features from classroom videos and behavioural logs. These features are combined with traditional statistical indicators and classified using SVM through a feature-level fusion strategy. Evaluated on simulated vocational classroom data, the fused model achieves 92.3% accuracy and an F1-score of 0.914, significantly outperforming standalone CNN-LSTM or SVM models. This model enables real-time, quantitative assessment of classroom engagement and supports timely teaching interventions.

Keywords: classroom participation; deep learning; support vector machine; SVM; feature fusion; vocational education.

DOI: 10.1504/IJICT.2026.151650

International Journal of Information and Communication Technology, 2026 Vol.27 No.9, pp.1 - 17

Received: 27 Aug 2025
Accepted: 19 Sep 2025
Published online: 11 Feb 2026 *

Title: The prediction model of higher vocational students' classroom participation based on the fusion of deep learning and support vector machine

Keep up-to-date