We’re looking for a post-doc! See a description of the position below.
ML algorithms have become important tools for leveraging healthcare data to improve patient outcomes and streamline hospital processes. Nevertheless, concerns regarding the reliability of these algorithms remain a barrier to their widespread adoption, as these algorithms may perform poorly in certain populations and/or decay in performance over time. The aim of this project is to develop tools that explain what the main causes for a performance gap. We’ll be drawing on various techniques, including causal inference, semi-parametric inference, and feature attribution methods. We’ll also be integrating these tools with sequential monitoring procedures to develop systems that both detect and explain performance drift over time. To test out these systems, we will be analyzing clinical ML algorithms that are slated for real-world deployment across EHR systems from three institutions: UCSF, ZSFG, and Duke University. At the end of the project, these tools will be distributed as an open-source software package for use by ML developers as well as hospital IT/data science teams.
This collaborative project includes team members with wide-ranging expertise:
- Julian Hong, a radiation oncologist who developed the SHIELD-RT algorithm for predicting acute care needs in radiation therapy patients and conducted one of the first RCTs for ML-guided care
- Romain Pirracchio, an anesthesiologist and biostatistician who has written a textbook on statistical modeling for ICU data
- Mi-ok Kim, a statistician who develops methods for causal inference, adaptive clinical trials, and longitudinal data
- Alexej Gossmann, a mathematical statistician at the FDA who, among other things, develops methods for biomedical imaging data
We are looking to hire a postdoctoral researcher to join the team. The position (100% funded) will be for 3-years. Salary and benefits are competitive.
The position requires a PhD degree or equivalent in (bio)statistics, computer science, or another relevant field. We are looking for someone who:
- has experience doing methods development
- can perform independent research
- can work and collaborate with a team
- has experience developing in a Linux environment, using git-based workflows, and writing in Python/R
If you are interested in this position, please submit the following materials:
- A cover letter
- A CV summarizing your education and work experience so far
- The names and email addresses of three references
- One representative publication
- A code sample
Please send these materials to [PI’s email] if you’re interested.
Screening of applicants will begin immediately and will continue as needed throughout the recruitment period. The fellow will start as soon as March 2023. The position will come with a competitive postdoc-level salary with great benefits for two years, with the ability to extend if things are going well.
The University of California is an Equal Opportunity/Affirmative Action Employer. All qualified applicants will receive consideration for employment without regard to race, color, religion, sex, sexual orientation, gender identity, national origin, disability, age, or protected veteran status.