With the development of digital information transmission and human-computer interaction methods, the conventional human-computer interface design approaches have failed to meet the demands for aesthetics, convenience and personalization in artistic design spaces. This paper constructs a deep reinforcement learning-driven adaptive interface optimization framework, integrates user behavior, visual semantics, layout structure and feedback evaluation for state modeling, designs multi-objective reward mechanisms, experience replay and target network update strategies, and introduces application implementation and feedback closed-loop update mechanisms. The experiments are conducted based on 5100 interface samples and 68,000 interaction sessions. The results show that the proposed method reduces the average task completion time to 11.8 seconds, increases the effective click rate to 92.6%, achieves a user satisfaction rate of 4.53, has a comprehensive performance index of 0.912, and maintains good stability even under high complexity and high concurrency conditions. It provides an expandable technical path for the intelligent optimization of artistic design interfaces and the integration of human-computer interaction.
Povzetek: This paper focuses on the optimization of the interactive interface in art design, and constructs a deep reinforcement learning framework that integrates visual semantics, layout structure, user behavior, and multiple objective rewards. It also includes experience summaries and feedback loop update mechanisms. After testing 5100 interface samples and 68,000 interaction sessions, the average task completion time was reduced to 11.8 seconds, the effective click-through rate reached 92.6%, the comprehensive performance value was 0.912, demonstrating excellent real-time and practicality.
