RL learns complex processes autonomously. No big data sets with the “right” answers are needed: the algorithms learn by experimenting. In 2017, RL algorithms beat the reigning Go World Champion.
The lecture intuitively shows “how” and “why” RL works. As a practical example, we apply RL to syphonic roof drainage: The choice of the “right” dimensions of trillions of possibilities is extremely difficult. Similar optimisation challenges appear in areas like dynamic pricing, game balancing, or player reward systems in non GamStop casinos. By the way, this prevents large buildings, such as airports or stadiums, from collapsing in heavy rain.
This talk shows how RL complements our existing machine learning solution, delivering a 56 per cent reduced fail rate.
Required audience experience
Basic familiarity with machine learning is helpful, but not required
Objective of the talk
• Develop an intuitive understanding of reinforcement learning
• Work with OpenAI gym
• Deploying a “bleeding edge” solution on 7,000 clients
