MIND2019_IM_RL_SR
Tutorials for MIND2019 on cognitive maps. Content includes:
1- Notebook applying Q-learning vs. random exploration to the taxi problem from open AI GYM.
2- Learning successor representations with SR (no replay). Source: Momennejad et al. 2017, NatHumBeh, Experiment 1
3- Learning successor representations with SR and replay (SR-Dyna). Compare to 2. Source: Momennejad et al. 2017, NatHumBeh, Experiment 1
Paper available here: https://www.momen-nejad.org/