Research on the Reinforcement Learning Strategy in the Crafts and Arts Design Automation Database System

This paper proposes a crafts and arts design automation database system based on reinforcement learning, which uses reinforcement learning algorithms to optimize design schemes and generate design works that meet the needs of artistic creation. By designing a database containing a variety of design elements (such as form, color, texture, etc.), the system can automatically learn the rules in historical design data and perform self-optimization through the Q-learning algorithm. In the experiment, this paper used a diverse data set containing traditional crafts and modern art designs. The system continuously adjusted the selection of design elements through the reinforcement learning model to maximize the creativity and aesthetics of the design. The experimental results show that the system using the reinforcement learning strategy can significantly improve the quality of design works, specifically, the design aesthetics score increased by 15.3%, the creativity score increased by 12.8%, and the creation efficiency was also increased by 20% compared with the traditional design method. In addition, the system also showed good optimization capabilities in terms of user acceptance and diversity, showing the great potential of reinforcement learning in the field of art design automation.