Artificial Intelligence

Lunch and Learn: An Introduction to Reinforcement Learning

Sunday, November 28, 2021
28
Nov

Recently, we hosted a Lunch and Learn session with Data Fusion Engineer and PhD Candidate, Jérémie Bannwarth. His discussion focused on providing an introduction to reinforcement learning: the area of machine learning focused on how intelligent machines (called “agents”) learn to interact with their environment to achieve a desired outcome. As Jérémie highlights, the agent’s behaviour is significantly influenced by its environment, as the rewards and punishments provided are dictated by the model. 


Deep reinforcement learning is a natural progression of the simple agent model - in this scenario, the agent is a deep neural network which allows its decision-making process to be much more complex in its relationship to the reward, providing more complex routes to outcomes.


Jérémie’s talk provides a great introduction to the key aspects of reinforcement learning, including how sparse and non-sparse reward models can impact the agent’s response and how punishments impact behaviour, along with a snapshot view of use cases and tools available for reinforcement learning projects and some examples of how it can be used. He also provides some great resources should you wish to continue reading about reinforcement learning, which we’ve linked below.


Resources to Learn More About Reinforcement Learning:


David Silver’s Lectures with Deepmind

OpenAI Gym and Stable Baselines 3 documentation

Reinforcement Learning by Sutton and Barto

Are you interested in hearing more about the business applications of AI technologies like Reinforcement Learning, Computer Vision, and Natural Language Understanding? Get in touch to find out more.


Click to Launch
L I T T L E    A L B E R T    D E M O
Want to stay in the loop? Subscribe to get updates from us.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
The Blog