Explainable Machine Learning Models for Healthcare AI

This tutorial extensively covers the definitions, nuances, challenges, and requirements for the design of interpretable and explainable machine learning models and systems in healthcare. We discuss many uses in which interpretable machine learning models are needed in healthcare and how they should be deployed. Additionally, we explore the landscape of recent advances to address the challenges model interpretability in healthcare and also describe how one would go about choosing the right interpretable machine learning algorithm for a given problem in healthcare.

Ankur Teredesai

Ankur M. Teredesai is the co-founder and Chief Technology Officer of KenSci. He also holds a Professorship in Computer Science & Systems at the University of Washington. Ankur's research spans data science with its applications for societal impact in healthcare. Apart from his academic appointments at RIT and the University of Washington, Teredesai has significant industry experience, having held various positions at C-DAC Pune, Microsoft Research, IBM T.J. Watson Labs, and a variety of technology startups. He has published more than 75 papers on machine learning, managed large teams of data scientists and engineers, and deployed data science solutions in healthcare. His recent applied research contributions include cost and risk prediction for readmission due to chronic conditions such as congestive heart failure. Other applications of his work have enabled predicting lengths of stay and sepsis as well as predicting medication pathways to lower risks of mortality and rehospitalization. He is Executive Director of the UW Center for Data Science, and serves as the Information Officer for ACM SIGKDD (Special Interest Group in Knowledge Discovery and Data Mining), the leading organization of industry and academic researchers in data science. He is currently an associate editor for ACM SIGKDD Explorations and IEEE Transactions on Big Data and serves on program committees of major international conferences in machine learning and healthcare.

Dr. Carly Eckert

Carly Eckert, M.D., M.P.H. is the Medical Director of Clinical Informatics at KenSci. In this role, Dr. Eckert leads and works with doctors, data scientists, and developers to identify patterns in patient data to predict risk that can cost-effectively improve care outcomes. Prior to her role at KenSci, Dr. Eckert was the Associate Medical Director for Catastrophic Care at the Department of Labor & Industries for the state of Washington. She trained in General Surgery at Vanderbilt University Medical Center and in Occupational & Environmental Medicine and Preventive Medicine at the University of Washington (UW).

She has also co-authored several publications on topics related to general surgery, occupational health, and occupational injury. She recently co-authored a publication accepted for presentation at AAAI: Death vs Data Science: Predicting End of Life. Dr. Eckert received her Masters of Public Health (M.P.H.) in Epidemiology from the University of Washington School of Public Health where she continues her studies as a doctoral student in the Epidemiology department. She received her Doctor of Medicine (M.D.) from the University of Oklahoma Health Sciences Center.

Muhammad Aurangzeb Ahmad

Muhammad Aurangzeb Ahmad is the Principal Data Scientist at KenSci. In this role, his work is focused on applying machine learning to solve problems within healthcare. His research at KenSci is focused on interpretable machine learning, fairness in machine learning, and causal machine learning models within the context of healthcare. Before coming to KenSci, Muhammad worked in applied machine learning in various domains, e.g., retail (Groupon), video gaming (Ninja Metrics), population studies (MPC), biomedical devices (Boston Scientific), and the energy sector (Con Edison). After working in different fields, Muhammad found his calling in healthcare, where he saw the great potential in using machine learning to improve the lives of people.

Muhammad holds a Ph.D. in Computer Science from the University of Minnesota-Twin Cities. He has taught machine learning and data science at the University of Washington – Tacoma, and he was a visiting research scientist at the Indian Institute of Technology at Kanpur. He has published more than 50 research papers on machine learning and data science.

Vikas Kumar

Vikas Kumar is a Data Scientist working at KenSci. In this role, Vikas works with a team of data scientists and clinicians to build consumable and trustable machine learning solutions for healthcare. His focus is in building explainable models in healthcare and application of recommendation systems in clinical settings. Prior to KenSci, Vikas was pursuing his doctorate in Computer Science at the University of Minnesota, Twin Cities.

Vikas holds a Ph.D. with a major in Computer Science and minor in Statistics from the University of Minnesota, Twin Cities. He has worked on modeling and application of recommendation systems in various domains, such as media, location, and healthcare. His focus has been to interpret the balance users seek between known (or familiarity) and unknown (or novel) items to build adaptive recommendations. Prior to his Ph.D., he completed his Bachelor's at the National Institute of Technology, India and worked as a software engineer in Microsoft India.