Publications

2026

1. 

   SMAC: Score-Matched Actor-Critics for Robust Offline-to-Online Transfer

   Nathan Samuel Lara and Florian Shkurti

   2026

2025

1. 

   STITCH-OPE: Trajectory Stitching with Guided Diffusion for Off-Policy Evaluation

   Hossein Goli, Michael Gimelfarb, Nathan Samuel Lara, and 3 more authors

   2025

2024

1. 

   Recurrent Policies Are Not Enough for Continual Reinforcement Learning

   Nathan Samuel Lara, Veronica Chelu, and Doina Precup

   In I Can’t Believe It’s Not Better Workshop: Failure Modes of Sequential Decision-Making in Practice (RLC 2024), 2024

2023

1. 

   Towards safe mechanical ventilation treatment using deep offline reinforcement learning

   Flemming Kondrup, Thomas Jiralerspong, Elaine Lau, and 5 more authors

   In Proceedings of the Thirty-Seventh AAAI Conference on Artificial Intelligence and Thirty-Fifth Conference on Innovative Applications of Artificial Intelligence and Thirteenth Symposium on Educational Advances in Artificial Intelligence, 2023

   Abs DOI

   Mechanical ventilation is a key form of life support for patients with pulmonary impairment. Healthcare workers are required to continuously adjust ventilator settings for each patient, a challenging and time consuming task. Hence, it would be beneficial to develop an automated decision support tool to optimize ventilation treatment. We present Deep-Vent, a Conservative Q-Learning (CQL) based offline Deep Reinforcement Learning (DRL) agent that learns to predict the optimal ventilator parameters for a patient to promote 90 day survival. We design a clinically relevant intermediate reward that encourages continuous improvement of the patient vitals as well as addresses the challenge of sparse reward in RL. We find that DeepVent recommends ventilation parameters within safe ranges, as outlined in recent clinical trials. The CQL algorithm offers additional safety by mitigating the overestimation of the value estimates of out-of-distribution states/actions. We evaluate our agent using Fitted Q Evaluation (FQE) and demonstrate that it outperforms physicians from the MIMIC-III dataset.