Hands-on tutorial on Deep Reinforcement Learning
jeu. 25 octobre à 19:00
Ever wanted to train a neural network to play video-games :) ? For our first event of the season we'll have Eustache Diemert from Criteo AI Lab giving a hands-on tutorial on modern approaches to Deep Reinforcement Learning. We will particularly emphasize practical tips to understand/debug the performance of the algorithms. It will consist in two parts: - Deep-RL Theory (30')
- Practical Session (45') For the practical part you are encouraged to come with your laptop. The exercises/demos will only require basic python knowledge as we'll use teaching material provided by the Criteo AI Lab researchers ( https://github.com/criteo-research/paiss_deeprl ) and provide access to a prepared environment. The practical part will be paced by regular case studies and polls so that people without a laptop can learn something and participate. Last but not least there will be drinks after the session so that people can socialize / ask follow-up questions etc :) Speaker Bio:
Eustache Diemert is a Staff Research Scientist at Criteo AI Lab ( http://ailab.criteo.com/ ), working on different research problems of the digital advertising industry such as click/sales prediction, bidding strategies, recommender systems and evaluation. Before that Eustache worked on solving different machine learning problems for social media and search engines companies.
Nous avons temporairement désactivé la possibilité de naviguer vers les tags.