Neural networks versus Logistic regression for patient readmission prediction
Brown Bag Lecture by Ahmed Allam, Ph.D. | 7/17/2018 11:00AM – 12PM | 7th Floor Conference Room, Bldg 38A
Research question: Can neural network models utilizing patients’ history in large administrative dataset outperform logistic regression for predicting 30 days all-cause readmission after discharge from a heart failure (HF) hospitalization event?
Using timelines of 272,778 patients, we (1) explored the systematic application of neural network-based models versus logistic regression for predicting 30 days all-cause readmission after discharge from an HF event, (2) examined the value of using patients’ history for boosting readmission prediction and (3) evaluated these predictions using a large administrative claims dataset. In this talk, we will report on the modeling approaches and the findings of this research, in addition to the implemented techniques for the interpretability of the models’ prediction.
Dr. Allam joined CEB as a Postdoctoral Fellow in November 2017. Previously, he was a Postdoctoral Associate at Yale University, working on biomedical informatics research projects in the Krauthammer Lab. He holds a PhD in Health Communication from University of Lugano, Switzerland, and Master of Science degrees in Computer Engineering from Politecnico di Milano and Politecnico di Torino, Italy.