Visit our website: http://bit.ly/2GtXaiw
Varoon Mathur, Technology Fellow, AI Now Institute
Abstract:
Electronic Health Records (EHRs) are increasingly being used for the training and deployment of machine learning systems in order to better health outcomes for patients. While there is significant promise due to the early success of deep learning diagnostic tools, there is likewise increasing concern around how such models might produce biased outcomes among patient groups, and what strategies might be employed to produce fairer machine learning algorithms. Audience members should walk away with a more clear understanding about how techniques around building fairer algorithms can be married together with the promise for better predictive tools for the future.
Add comment