FrozenLake environments

该数据集包含了一系列用于评估匹兹堡学习分类系统在强化学习任务中性能的环境。这些环境通过网格大小（M）和随机性（p_slip）进行参数化，从而产生了随着复杂性增加的各种组合。基于不同的网格大小和随机性水平，数据集共包含了12个环境，其任务专注于强化学习。

This dataset includes a series of environments designed to evaluate the performance of Pittsburgh Learning Classifier Systems in reinforcement learning tasks. These environments are parameterized by grid size (M) and stochasticity (p_slip), generating various combinations with increasing complexity. The dataset contains a total of 12 environments across varying grid sizes and stochasticity levels, all focused on reinforcement learning tasks.