Dear students,
We slightly updated the code of the simple maze coding exercise. If you haven’t started working on it yet, please download the new version and you can ignore the paragraph below.
In the previous version, you had to code a little bit more than the space we provided for you. To be precise:
- You had to update the current action with the next action (because SARSA is an on-policy method)
- You had to set the matrix provided for the eligibility traces to 0 after each episode.
These parts are implemented in the new version and we will release the solutions next week.
Meanwhile, we encourage you to use Moodle actively for your questions :)
Have a nice evening and take care,
your TA team
