This page is located in archive.

Open AI Gym

Another set of simulation-based topics can be defined on top of the AI Gym environment, implemented in Python.

We have not succeed to find a reliable way how to install this library on Windows. (If you know how to do that, let us know!) On Linux machines, the library works fine!


The library is primarily aimed at testing and comparing reinforcement learning algorithms. In individual tasks you should find certain control strategy. This can be formulated as a search in the space of possible control strategies, and solve it using an evolutionary algorithm. You should pay attention to the representation of those strategies. For some environments, the task can be simplified e.g. by fixing the starting state, and search for the optimal control sequence just for this starting state.

If you want to choose this topic, inform us ASAP, we have to discuss the details. We have to approve the specific topic in advance!

courses/a0m33eoa/semestral_tasks/aigym/start.txt · Last modified: 2021/09/20 12:21 by xposik