Minimizing confounding is a key challenge to ensuring the fidelity of observational assessments of the real-world safety and effectiveness of medical products. Significant advances have been made in leveraging data-driven machine learning approaches to efficiently reduce potential confounding. This webinar will focus on super learning and targeted maximum likelihood estimation, in particular, as solutions to reducing bias in observational studies of electronic health record data.
![](https://i.ytimg.com/vi/CoO0Y54QxdM/maxresdefault.jpg)