Yaniv Romano, Associate Professor

Prof. Romano's research develops statistical foundations for making AI reliable and useful in the real world. He is focusing on uncertainty quantification, out-of-distribution detection, robust prediction under changing conditions, reproducible AI-powered scientific discovery, and the principled use of synthetic data to overcome sample-size limitations.

My work develops tools for hypothesis testing and online statistical inference, as well as new machine-learning paradigms, all aimed at helping researchers reliably extract what the data can genuinely reveal. These challenges are central to cancer research, where researchers must integrate complex imaging, molecular, and clinical data; identify reproducible biomarkers and treatment-response signals; and reliably predict outcomes such as survival and disease progression.