Thibault Marzullo

About

Thibault Marzullo is a rising researcher at the forefront of intelligent building energy management, specializing in the intersection of reinforcement learning (RL) and inverse reinforcement learning (IRL). His work addresses the growing complexity of building energy systems by pioneering control approaches that enable autonomous, efficient decision-making. Marzullo’s most impactful contribution, “Reinforcement learning building control approach harnessing imitation learning” (2023, 54 citations), demonstrates how RL can be enhanced by imitation learning to optimize energy use, drawing inspiration from successes in autonomous vehicles and robotics. He further advances the field with “Inverse reinforcement learning control for building energy management” (2023, 26 citations), where he develops methods for learning optimal control policies by observing expert behavior, reducing the need for manual reward design. This dual focus on RL and IRL positions Marzullo as a key innovator in making buildings smarter and more sustainable. His work, though recent, has already garnered significant attention, reflecting its potential to transform energy management in increasingly complex built environments.

Research Focus

Key Achievements

Top Papers

1. 1
   Reinforcement learning building control approach harnessing imitation learning

   54 citations · 2023
2. 2
   Inverse reinforcement learning control for building energy management

   26 citations · 2023
3. 3
   Inverse Reinforcement Learning Control for Building Energy Management

   3 citations · 2023

Key Collaborators

Contact & Links