My caption 😄

Towards Real-world Reinforcement Learning for Building Control

The project aims to develop enabling methods for practical deployment of reinforcement learning for building control, such as:

  • Initialize a policy with historical data through imitation learning
  • Estimate a policy’s performance without running it on the actual system via off-policy evaluation
  • Learn on the real buildings with limited samples through model-based RL.
Mario Bergés
Mario Bergés
Professor of Civil and Environmental Engineering

My research interests vary, but generally gravitate towards the development of technologies to make our built enviornment and the communities in them more autonomous and efficient.