The Reinforcement Learning and Robustness Lab (RLsquared) conducts research on models and algorithms for reinforcement learning and data-driven decision making.

We want to deliver real-world impacts through reinforcement learning by designing methods that practitioners can use out-of-the-box.

We are current working on these topics:

• Robust reinforcement learning and optimization
• Exploration in reinforcement learning
• Fast mathematical optimization
• Interpretable reinforcement learning
• Value function and model approximation techniques

Our current applications are:

• Management of invasive species with Jenica Allen
• Data-driven pest management in agriculture with Shadi Atallah
• Machine learning for magnetosphere with Matthew Argall
• And others