Date: Tuesday, February 20, 2018

Time: 03:00pm – 04:15pm

Location:  Klaus Advanced Computing Building, Room 2443

Abstract:  How can we distill multiple aspects of raw and noisy electronic health record data, such as diagnoses, medications and procedures, to a few concise clinical states without human-annotated labels? In this talk, we present a review of works tackling this challenge through the use of tensor factorization methods. Several aspects of this problem will be discussed such as scalable and efficient computations, interpretability of the results and handling temporally-evolving data.

Bio: Ioakeim (Kimis) Perros earned the Diploma and M. Sc. degrees in Electronic & Computer Engineering from the Technical University of Crete, Greece, in 2012 and 2014 respectively. He is currently with the SunLab group, working as a Research Assistant and pursuing a Ph.D. degree in Computational Science & Engineering from the Georgia Institute of Technology. He has interned at the Health Informatics Division of Weill Cornell Medicine, the RD&D Department of Sutter Health and the Healthcare group of Microsoft Research, Cambridge. He has published in top Data Mining (KDD, SDM, ICDM) and co-authored papers in top Machine Learning (NIPS) and High-Performance Computing (SC, IPDPS) venues. His current research focus is on developing tensor methods for Computational Medicine.