Treatment Effect Modelling and Recommendation

We aim to leverage machine learning to estimate treatment effects from observational clinical data, enabling personalised therapeutic recommendations. Key challenges include handling artefacts in clinical measurements, capturing complex temporal dynamics, and the lack of robust evaluation metrics. Our innovations focus on optimising treatment policies, enhancing clinical decision support, and ultimately improving patient outcomes through data-driven precision medicine.

Publications

Luo, Zhiyao, et al. "DTR-bench: an in silico environment and benchmark platform for reinforcement learning based dynamic treatment regime." arXiv preprint arXiv:2405.18610 (2024). paper
Luo, Zhiyao, et al. "Position: reinforcement learning in dynamic treatment regimes needs critical reexamination." Proceedings of the 41st International Conference on Machine Learning. 2024. paper, code
Luo, Zhiyao, Peter Watkinson, and Tingting Zhu. "NurSpecialist: Duel-Agent Reinforcement Learning for Dynamic Hospitalised Intervention Regimes using Electronic Health Records." (2022). paper