RL Odyssey 3: Representation Learning for RL

Learning from visual observations is a fundamental yet challenging problem in Reinforcement Learning (RL). Although algorithmic advances combined with convolutional neural networks have proved to be a recipe for success, current methods are still lacking on two fronts: (a) data-efficiency of learning and (b) generalization to new environments. This talk will introduce representation learning methods that help improve the performance of RL agents. Some of the topics covered will be Data augmentations, learning invariant and temporal representations, and contrastive self-supervised learning.

L'evento inizia alle 18:00, il talk inizia alle 18:30.

Speaker: Ameya Pore
Aula: T04

REGISTRAZIONE:
https://www.eventbrite.it/e/888311352167