Award
National Library of Medicine 5R01LM013614-04
Semi-supervised Approaches to Denoising Electronic Health Records Data for Risk Prediction
Recipient
Harvard University D/B/A Harvard School of Public Health, Boston, MA, United States
Award Amount
$334,713.00
Ceiling
$334,713.00
Awarded
April 15, 2024
Identifier
5R01LM013614-04
Development of advanced statistical and machine learning methods for accurate risk prediction using noisy and heterogeneous electronic health record data, facilitating personalized disease management.
Description
This project aims to develop semi-supervised risk prediction methods leveraging noisy surrogates and limited human annotations, transfer learning techniques to adapt models across different EHR systems, and distributed learning approaches to handle high-dimensional data while respecting patient privacy.