Lecture | Tuesday Talks: Science Insights
Teaching Machines to Learn
- Date
- Tuesday 8 April 2025
- Time
- Address
-
Gorlaeus Building
Einsteinweg 55
2333 CC Leiden - Room
- CM1.26
In early March the 2024 Turing Award was officially announced and awarded to Andrew Barto and Richard Sutton for their groundbreaking work in developing the conceptual and algorithmic foundations of Reinforcement Learning (RL).
Join us for this Tuesday Talk, where Aske Plaat will talk about Reinforcement Learning and why Barto and Sutton’s contributions were awarded the most prestigious award in our discipline.
Join this edition of Tuesday Talks: Science Insights!
Everyone who is interested in this edition's topic is welcome. Please let us know if you are attending.
I will joinProgramme
The programme of the Tuesday Talks: Science Insights is:
15.30 hrs | Take a seat |
16.00 hrs | Tuesday Talks: Science Insights |
17.00 hrs | Drinks |
Inspired by animal learning
Reinforcement Learning (RL) is learning by trial and error. An agent chooses an action, performs it in the world, and learns from the feedback it receives. Inspired by how humans and animals learn from experience, RL mimics how a child learns to walk through repeated attempts and adjustments, or how a dog learns tricks through rewards. Computers are no different—they can be trained to make decisions and solve problems using this method.
From games to real-world impact
The foundational work of Barto and Sutton has paved the way for modern RL techniques. As a general learning method, RL has achieved remarkable success in areas such as game playing, robotic locomotion, and real-world applications in health and finance. Understanding these principles offers valuable insights into how machines can be taught to tackle complex challenges.
About the speaker
Aske Plaat is a professor in artificial intelligence at our faculty and heads the Sustainable Energy Learning Lab and the Reinforcement Learning Group. He is the author of two textbooks on Deep Reinforcement Learning, and his research focuses on reinforcement learning and self-learning systems. Among his many achievements, he has won the Novag Award for the best computer chess publication and co-authored an ERC Advanced grant on combinatorial algorithms for high-energy physics.
About Tuesday Talks: Science Insights
Tuesday Talks: Science Insights is Leiden Science’s monthly dose of research inspiration from our staff, for our staff and all others curious minds. Every second Tuesday of the month from 16:00 hrs till 17:00 hrs, drinks and meet up afterwards till 18.00 hrs.