All Times EDT
Machine learning (ML) has been used to support regulatory decision making in post-market drug safety surveillance. In this presentation, we provide a landscape of ML applications within the Division of Biometrics VII of Center for Drug Evaluation and Research, in Food and Drug Administration. We present the present utility and potential of using ML for regulatory decision making. Additionally, practical challenges and considerations using ML in solving prediction and causal inference problems will be discussed. Specifically, we highlight the reproducibility and transparency issues of using ML, challenges of using electronic health record and claims data in ML and natural language processing applications, interpretability issue of applying ML methods in causal inference, and more.