Join our webinar on the use of machine learning to construct predictive models for outcomes in healthcare from data about patients and their encounters with providers. When deployed, the predictive models provide patient-specific predictions about the likelihoods of outcomes of interest, such as the probability that a patient being discharged will be readmitted within a short time horizon. Speaker Eric Horvitz, distinguished scientist will focus on the challenge of predicting readmissions as a representative challenge in prediction and decision making. After describing the construction of models for predicting readmissions he will demonstrate how the predictive models can be used within cost-benefit analyses that provide guidance on practice and policy. Learn more about the speaker at: research.microsoft.com
Video Rating: 5 / 5