AI Advances and Applications Virtual Seminar – Dr. Mattia Prosperi
- This event has passed.
AI Advances and Applications Virtual Seminar – Dr. Mattia Prosperi
September 16, 2020 @ 12:00 pm - 1:00 pm
AI Advances and Applications Virtual Seminar Series
“The Big Short with AI in Biomedical Sciences: When Actions Don’t Follow Predictions” by Dr. Mattia Prosperi
Associate Professor, Preeminence Faculty, Dept. of Epidemiology, Data Intelligence Systems Lab (DISL)
Wednesday, September 16, 2020
ABSTRACT:
Artificial intelligence (AI) is increasingly becoming key to precision medicine –from identifying disease risks and taking preventive measures, to making diagnoses and personalizing treatment for individuals. Precision medicine, however, is not only about predicting risks and outcomes, but also about weighing interventions. Interventional clinical predictive models need to calculate the so-called counterfactuals, i.e. alternative scenarios (changing variables that would not be naturally modified), requiring the correct specification of cause and effect. In biomedical research, observational studies are commonly affected by bias, such as confounding, selection, and indication. Without robust assumptions, often requiring a priori domain knowledge, derivation of causal effects can be flawed. Data-driven models might have very high accuracy, even in multiple test sets, yet be non-causal and produce the wrong predictions when used to evaluate counterfactuals. Therefore, the premise that data-driven prediction models lead to trustable decisions/interventions for precision medicine is questionable. When pursuing intervention modelling, the bio-health informatics community needs to employ causal AI approaches.
Bio:
Dr. Mattia Prosperi is Associate Professor of Preeminence at UF and leads the Data Intelligence System Lab in the College of Public Health and Health Professions. He holds a Master of Engineering and a PhD in Computer Science, and was faculty at University of Manchester, UK, before moving to UF. His research includes development of theory and applications in artificial intelligence tailored to biomedical sciences, in particular precision medicine and precision public health.