Actions and Detail Panel
Data Science Tech Talk @ Campus London
Tue 20 September 2016, 19:00 – 21:00 BST
Join us for a series of Tech Talk's exploring state-of-the-art techniques and their practical applications. We'll be covering a wide range topics such as Machine Learning, Reinforcement Learning, Computer Vision, Data Visualisation and more.
Title: Introducing Deep Reinforcement Learning
Abstract: Reinforcement learning is a powerful general framework that addresses a wider scope of control tasks than more conventional supervised learning methods. However, mathematically sound algorithms for reliably learning optimal control policies, such as Q-learning, tend to place heavy demands on storage and hence do not offer viable solutions to many real-world problems. This talk will provide a high-level overview of how to leverage deep neural networks to make Q-learning feasible for large (even infinite) state spaces. I will also discuss popular real-world examples such as DQN (for playing Atari games).
Speaker: Petar Veličković
Petar is a Research Assistant in Computational Biology within the Artificial Intelligence Group of the Computer Laboratory.
If you are interested in progressing further with Machine Learning, learn more at the Data Science Bootcamp in Python.