For the bootstrapping method, we propose using diversity sampling to improve the solution diversity, auxiliary targets to enhance the modelsвЂ™ quality and Gaussian mixture models to improve the sample quality. We will propose some modifications to either adapt the methods to the task or improve their efficiency and performance. The methods will be bootstrapping conditional generative models, controllable & uncontrollable procedural content generation via reinforcement learning (PCGRL) and generative playing networks. This paperвЂ™s goal is to conduct an experimental study on four recent deep learning procedural level generation methods for Sokoban (size = 7 Г— 7) to explore their strengths and weaknesses and provide insights for possible research directions. Overall, PCGRL shows superior quality and diversity while generative adversarial networks exhibit the least control confusion when trained with diversity sampling and auxiliary targets.Дeep learning for procedural level generation has been explored in many recent works, however, experimental comparisons with previous works are either nonexistent or limited to the works they extend upon. On average, auxiliary targets increases the quality by 24% and sampling conditions from Gaussian mixture models increases the sample quality by 13%.

The results show that diversity sampling at least doubles the unique plan count in the generated levels. For the bootstrapping method, we propose using diversity sampling to improve the solution diversity, auxiliary targets to enhance the models' quality and Gaussian mixture models to improve the sample quality. This paper's goal is to conduct an experimental study on four recent deep learning procedural level generators for Sokoban to explore their strengths and weaknesses. Finally, a study was conducted on the performance of our algorithm to generate levels of three different game experiences, from which we demonstrate the ability of IORand to satisfactorily and consistently solve the generation of levels that provide specific game experiences.Дeep learning for procedural level generation has been explored in many recent works, however, experimental comparisons with previous works are rare and usually limited to the work they extend upon. Then, the design of our hybrid PCG algorithm, IORand, whose reward function is based on the proposed level evaluation method, is presented. Moreover, the design of a new method for evaluating video game levels is presented, aimed at evaluating game experiences, based on graphs, which allows identifying the type of interaction that the player will have with the level. Our study includes a presentation of current PCG techniques and why a hybridization of approaches has become a new trend with promising results in the area. In this work we present the intelligent orchestrator of random generators (IORand), a hybrid procedural content generation (PCG) algorithm, driven by game experience, based on reinforcement learning and semi-random content generation methods.

The talos principle higher ground