Abstract: Machine learning can be challenging for conventional information technology resourcing in healthcare, especially the requirements of more intense computing, an unfamiliar technology toolkit, rapid interactions with data stores, and the need for associated processes of data governance. In this abstract we describe the process flow for machine learning-based health data science projects, the extension of Duke Health’s protected workspace solution learning, and the collaboration of resources and expertise between our university and health system.
Learning Objective 1: 1. After participating in this session, the learner should be better able to: assess the requirements of machine learning and articulate how conventional information technology can be challenged.
Learning Objective 2 (Optional): 2. After participating in this session, the learner should be better able to: describe the design of a protected analytics solution and its extension for machine learning.
Shelley Rusincovitch (Presenter)
Charles Kneifel, Duke University
Michael Gao, Duke University
Billy Willis, Duke Health
James Fayson, Duke Health
Stephen Blackwelder, Duke Health
Lawrence Carin, Duke University
Ricardo Henao, Duke University