Research

Kan Chen develops statistical and machine learning methods for reliable causal reasoning in complex real-world data. A central theme of his research is the development of theoretically principled, interpretable, robust, and generalizable methods for scientific and clinical decision-making.

Methodologically, his work spans causal inference, causal machine learning and AI, high-dimensional statistics, and large-scale biomedical data analysis. His research particularly focuses on unmeasured confounding, selection bias, residual confounding, interpretability, robustness, and generalizability.

Substantively, his work is motivated by applications in clinical trials, oncology, genomics and epigenomics, patient-centered health research, and policy evaluation. The long-term goal of his research program is to develop reliable causal AI tools that improve human health, support scientific discovery, and inform evidence-based policy.