<ul><li>Serious Games are incorporating procedural content generation (PCG) to enhance player experience.</li><li>A new study introduces a framework for automated evaluation of PCG integration in Serious Games.</li><li>The framework utilizes deep reinforcement learning (DRL) game testing agents to assess different versions of PCG in a card game-based SG.</li><li>Results show that DRL agents trained on genetic algorithm-based PCG versions outperform those trained on random NPC creation in terms of win rate and training time.</li></ul>

A modular framework for automated evaluation of procedural content generation in serious games with deep reinforcement learning agents

Discover more