A periodic meeting to discuss about Machine/Deep Learning topics.

Motivated by the presentation... focused on the discussion!

If I have an idea and you have an idea, we both have an idea. But, what happen if we share it?... We have TWO ideas!! Wow

Next session

To attend, please send an email to: [email protected]

22/02/2018 -Challenge on Deep Reinforcement Learning (DRL)

> Room D+i, 11-13h, ground floor CITEXVI building, Campus Universitario de Vigo (CUVI)

The aim of this practical session is to solve in groups a proposed RL problem in 2 hours approx. Are you ready for the challenge!! (prizes are expected and sponsors are more than welcome!).

More details will be available soon!. By Pablo Dago and David Jiménez

Pablo Dago - Multimodal	Adrián Portabales
Introduction to Reinforcement Learning	Deep Reinforcement Learning - Advanced details
In this part we will review the fundamentals of Reinforcement Learning: principles, advantages, disadvantages, limitations and possible use cases. For this, a theoretical-practical approach will be carried out that will conclude with practical examples of application of this paradigm through "notebooks" and possible use cases.	Thanks to his experience in the use of reinforcemente learning, Adrián Portabales will show us a set of "tips & tricks" about how to successfully use "Deep Reinforcement Learning" in real use cases.

Tentative Future Sessions

We are planning the following sessions (dates could be modified):

• March, 22: "The dark side of Deep Learning: adversarial attacks and how to defeat them. In this workshop we will talk about how to generate adversarial examples that drastically fool deep learning architectures and different alternatives to reduce its influence."

We encourage everybody to make a proposal of presentation. Just send an email to: [email protected]