Enhancing Student Engagement in Smart Classrooms Using Reinforcement Learning Algorithms

• Published in: Cross Strait Quad-Regional Radio Science and Wireless Technology Conference pp. 1 - 4
• Main Authors: Yantao, Li, Wei, Ji
• Format: Conference Proceeding
• Language: English
• Published: IEEE 04.11.2024
• Subjects: Accuracy; Deep learning; Educational technology; Heuristic algorithms; Long short term memory; personalized learning; Q-learning; Q-learning algorithm; Real-time systems; smart classroom; student engagement; teaching strategy optimization; Wireless communication
• ISSN: 2377-8512
• DOI: 10.1109/CSRSWTC64338.2024.10811575

This study aims to dynamically optimize the teaching strategy in smart classrooms by applying the Q-learning reinforcement learning algorithm to improve students' classroom participation. This paper designs a complete teaching feedback system that can collect students' classroom behavior data in real time and dynamically adjust the teaching strategy based on the Q-learning algorithm to improve the teaching effect. The experiment verifies the application effect of the Q-learning algorithm in smart classrooms by simulating three different teaching scenarios: large-class teaching, group cooperation, and personalized tutoring. The experimental results show that the Q-learning algorithm can effectively improve students' participation, especially in the group discussion and personalized tutoring modes, where the students' participation rate increases from 50% in the initial stage to 85%. In contrast, although the participation rate of the lecture mode increases less, it maintains a relatively stable growth. The research results further prove the effectiveness of the Q-learning algorithm in optimizing teaching strategies in different classroom scenarios and demonstrate its potential for wide application in smart education systems. In addition, this paper discusses the limitations of the study and possible future research directions, and points out that more deep learning algorithms, such as deep Q networks (DQN) or long short-term memory networks, can be introduced to enhance the intelligence level of the system and further optimize the personalized learning path.