Reform and Innovation of Civic and Political Education Based on Deep Enhanced Learning Methods

Abstract

The use of an initial state value function and an optimal strategy are used in this paper to solve educational problems based on deep reinforcement learning. Deep reinforcement learning’s approximate function is defined, and the matrix model is created by training tuning using learning methods like gradient descent. To analyze the modeling process of reinforcement learning, reward values are added to the Markov decision transfer matrix and the expected value of cumulative returns is calculated. The weights are trained using the Bellman equation to enhance the algorithm’s stability. In evaluating the effect of reform and innovation in Civic Education, the teacher education concept is rated as 10 points. The reform and innovation of civic education, combined with deep reinforcement learning, can promote the reform of education and teaching modes, improving the efficiency and quality of education.