Obstacles and Fire States MDP

该数据集呈现为一个N×N的网格世界，其中包含障碍物和火焰状态，旨在评估强化学习算法的性能。该数据集包含了多种不同实例的网格世界，每个实例中的障碍物和火焰状态各不相同，以便对算法的表现进行全面的评估。每个实例的规模为N×N（具体大小因实例而异），所涉及的任务是解决马尔可夫决策过程（MDP）问题。

This dataset comprises N×N grid worlds with obstacles and fire states, aiming to evaluate the performance of reinforcement learning algorithms. It includes multiple distinct grid world instances, each featuring unique obstacle and fire configurations to enable comprehensive performance evaluations of algorithms. Each instance has a variable N×N scale, with the exact size differing across instances, and the core task for each instance is to solve Markov Decision Process (MDP) problems.