- Táto udalosť už prebehla.
MLMU BA: Reinforcement learning 1: deep Q networks 2020
18. marca 2020 @ 19:00 - 21:00
Save the date, another MLMU meetup in Bratislava incoming! We will meet at Progress Bar on Wednesday, 18.3.2020, 19:00.
Please help us manage capacity by registering via Eventbrite:
Exceptionally, talk will be in Slovak (slides in ENG) followed by discussion and networking with refreshments at the venue in ProgressBar. If you can’t join us, don’t be sad, our talk will be livestreamed via facebook and link shared later on.
Event: Reinforcement learning 1 – deep Q networks
Sylabus of the talk:
1. Reinforcement learning versus supervised learning
2. Deep Q Network (DQN) – why it’s naive NN not working
3. Advanced architectures: noisy DQN, dueling DQN, rainbow DQN, attention DQN and hands-on practices how to create network architecture
4. Real examples (Atari, Super Mario, Doom 2) – how is network interacting with given world.
Whole talk is full of real examples of pytorch integration and results. Reinforcement learning is a machine learning method, which is trying to find optimal strategy (trajectory) in Markov decision process (in this case, game). Agent (usually deep neural network) is actively interacting with environment, considering current status to choose actions (policy) as output, evaluing rewards and punishments. Goal is to maximize sum of rewards. Agent does not know what is right behaviour, but has knowledge of actions and rewards. Based on this, reinforcement learning is capable of finding better strategy than human.
Michal Chovanec is a researcher in AI field. Currently is working in Tachyum in Bratislava, developing AI accelerator of new processor. His PhD degree is from University in Žilina. His favorite topics are reinforcement learning, robotics (winner of Gold Medal in Istrobot challenge) and modelling of red blood cell in research group Cell in fluid (FRI, UNIZA). His hobbies are hiking, archery and martial arts.
Zmena programu vyhradená. Uvedené údaje sú len informačné.
Stránka Rýchle.sk nie je organizátorom podujatí.
Obrázky podujatí môžu podliehať autorským právam.