Mark Fearnley is a Machine Learning Architect at Atos – a global leader in digital transformation. Working from the London based Atos AI Lab(under the Atos and Google Cloud’s global partnership), his days are spent creating cloud-based Machine Learning pipelines, and teaching people how to gain practical benefits from AI. He specialises in Deep Supervised and Reinforcement Learning, backed by a history of working with Web and Mobile development, Containerisation and DevOps. www.atos.net
Talk: Seeking Rewards: The Future of Reinforcement Learning in Industry
The use of AI and Machine Learning has massively grown in popularity over the last few years. We see it on a daily basis, with search, recommendations and photo apps being among the many products powered by modern advancements in Deep Learning. Using vast quantities of historic data, we can train machines to mimic intelligence by learning from examples of right and wrong – but can we use a more natural way of teaching machines?
Reinforcement Learning takes inspiration from nature – learning by interacting with an environment, trial and error and maximising rewards. In this talk, Mark will explain what Reinforcement Learning is, where it can be used and how it differs from the Machine Learning approaches commonly used nowadays.
Examples of this kind of AI can be seen in the wild, with AlphaGo Zero and OpenAI 5 showing how this technology can be used to train machines to outperform top-level professionals in massively complex games. Despite showing such promising results on a global stage, we don’t see wider adoption of these techniques. Mark will talk about why that is, why we will see it’s use grow rapidly and he will showcase some practical examples of the power of Reinforcement Learning in action.