Generation and Optimization of Teaching Decision Generation Under a Smart Teaching Environment

Abstract

Teaching decisions need to be optimized to guide learners to stick to the right learning method, and maintain learning interests. However, the existing evaluation systems for college teachers’ teaching decision ability cannot adapt to online teaching decision-making. What is worse, the previous studies on college teachers’ teaching decisions rarely consider online teaching decisionmaking. Therefore, this paper attempts to optimize college teachers’ teaching decision under the smart teaching environment. Specifically, the roadmap of teaching decision generation and optimization was presented, the teaching decision bases were specified for different teaching decision scenarios, and a teaching decision model was established under the smart teaching environment. In addition, teaching decisions were generated based on deep reinforcement learning algorithm, and optimized by certain rules under the experience replay mechanism. The proposed algorithm was proved effective and feasible through experiments.