- How to train AI to play Computer Games? 101
Did my computer just beat me? How can a computer learn in just one hour to be that good?! There are many different ways to learn certain skills. You can learn from example (supervised learning), you can seek certain patterns and find correlations between events (unsupervised learning) or you can learn through your own actions (reinforcement learning). We humans learn a lot through our own actions, but how could a computer do that? Join us to learn about the basics on how computers can learn this way! In this reinforcement learning 101 meetup we will cover the basic about reinforcement learning and use Deep Q Learning and Policy Gradient to solve certain problems. Practical information: the Meetup is held in the central hall of LabS55 at Lulofsstraat 55. It's a ten minute walk from Leeghwaterplein and a 15 minute walk from Station Holland Spoor and Haagse Hogeschool. 19:00 - 19:30 : Walk in and have something to drink. 19:30 - 20:00: First part of lecture/discussion. 20:00 - 20:15: Break. 20:15 - 20:45: Second part of lecture/discussion. 20:45 - 21:00: Hang out, drinks and discuss AI related topics. There is parking space in the street in front of the location. There is also free parking space in the street behind the fences. You are permitted to park here (it says you're not permitted, but I will notify security beforehand). If both are full, you can also park at Gamma, which is a one minute walk from the location.