Application of Deep Reinforcement Learning Algorithm for Adaptive Learning Path Planning in Immersive Learning Environment

Current traditional learning path planning methods are difficult to meet the needs of personalized learning, especially in immersive learning environments, where learners face problems such as cognitive burden and low learning efficiency. This paper aims to use deep reinforcement learning algorithms to achieve adaptive learning path planning. First, the state space of the immersive learning environment is constructed and the state variables are defined, including the learner’s knowledge level, learning preference, task difficulty, and real-time feedback information. Then, a deep Q network (DQN) algorithm is designed, which regards the learning path as an action choice and evaluates the learner’s learning progress and achievements by setting a reward function. The reward function comprehensively considers the learner’s learning time and task completion status, uses the experience replay mechanism to store past state-action-reward information, and updates the strategy through the target network technology to optimize the learning path selection. In learning path planning, the task completion rate of deep reinforcement learning is close to 98%. The deep reinforcement learning algorithm effectively overcomes many limitations of traditional learning path planning in an immersive learning environment and opens up a new path for personalized learning.