采用增强学习算法的排课模型

Arranging model of university timetable based on reinforcement learning

时间表问题是典型的组合优化和不确定性调度问题。课表问题是时间表问题的一种形式。分析了排课 问题的数学模型,并研究了用增强学习(Reinforcement Learning)算法中的Q学习(Q-Learning)算法和神经网络 技术结合解决大学课表编排问题,给出了一个基于该算法的排课模型,并对其排课效果进行了分析和探讨。

The University timetable arranging problem is an important task for Academic Affairs Office. Our previous work on formulating this task for solution by the reinforcement learning algorithm Q-learning is summarized. The mathematics model of curriculum arrangement is analyzed and how to extend the BP neural network architecture to apply it to estimate the value of timetable status is shown. Results are presented applying this approach to a gym course arranging based on reinforcement learning.