Timezone: »

Multi-group Reinforcement Learning for Electrolyte Repletion
Promise Ekpo · Barbara Engelhardt

Most off-policy reinforcement learning methods for specifying treatment policies in EHR data have a heterogenous patient population as well as different complications that are generally not considered in identifying optimal treatment policies as patient subgroups are hard to model. In this work, we use multi-group Gaussian process regression in a fitted Q-iteration framework to model diverse patient subgroups and adapt the optimal policies in a personalized manner as we approximate these functions across the full patient population. We apply our multi-group reinforcement learning (MGRL) model in specifying optimal treatment policies in recommending electrolyte repletion to ICU patients with several comorbidities in different ethnic groups. When utilized in clinical settings, we show that these policies learn interpretable differences in the datasets for the distinct patient subgroups.

Author Information

Promise Ekpo (Princeton University)

Promise Ekpo Osaine research interest is focused on the application of Machine learning in Healthcare especially in sequential decision processes.   A first-class (Summa Cum Laude) graduate of the Department of Computer Engineering, Class of 2020, University of Benin, Promise was the best graduating student of her department (CGPA: 4.75/5.00). She holds several awards such as Shell Undergraduate Scholarship 2016, NNPC Quiz (SPE Conference, 2018), Huawei and Glo Students Award for Academic Excellence 2019, and Michael Taiwo Scholarship 2020.   Promise was an intern with Shell where she developed technical skills in SQL, Python, and data management applications such as EP Catalogue, SharePoint, and ASSAI. She was involved in the Shell Eco-Marathon Monitoring Global project, where she did sentiment analysis and reporting. She has also worked with Union Bank as a data engineer where she did Extract, Transform, and Load processing using SQL Server Management tools and Visual Studio Code.   At the University of Benin, she held leadership roles such as; Secretary-General of Computer Engineering Association, Secretary-General of Engineering Final year committee, Academic coordinator Of Gospel pillars Campus fellowship, University Coordinator of the U-Genius Club for student academic outreach and free mentorship.   Promise owns a mentorship platform aimed at improving the educational opportunities of students and is the convener of the “Get an Undergraduate a scholarship”, project She plays piano for community church as well as campus fellowship.

Barbara Engelhardt (Princeton University)

More from the Same Authors