Manuscript received August 21, 2024; revised September 10, 2024; accepted October 16, 2024; published January 9, 2025
Abstract—Traditional approaches often fall short in effectively teaching complex subjects like probability and dynamic programming, especially in contexts requiring high engagement and individualized learning paths. This paper presents a simulation-based experimental study exploring the potential of an Artificial Intelligence (AI) adaptive learning system through the development of a game-based learning tool. The system utilizes dynamic programming principles and decision tree regressors to adjust the game complexity in real-time, based on simulated student performance. The adaptive dice game provides personalized learning experiences that improve both engagement and comprehension of key mathematical concepts. The experiment evaluates how adaptive difficulty settings influence strategic decision-making and learning outcomes. The results demonstrate that adaptive learning systems can significantly enhance mathematical education by offering customized learning paths that improve understanding of complex concepts. This study contributes to the discussion on the potential of integrating AI with educational technologies to enhance learning outcomes, particularly in disciplines that demand high analytical skills.
Keywords—game-based learning, adaptive learning, Artificial Intelligence (AI), probability
Cite: Xiao Xu, "Game-Based Adaptive Learning in Probability Education," International Journal of Information and Education Technology, vol. 15, no. 1, pp. 1-7, 2025.