[WINS] CS Seminar Thursday, Sept 29, 11:15am to 12:15pm * ISC 3300*
Wei Ding
SEMINAR TOMORROW AT THE ISC, 3RD FLOOR, ROOM 3300 (NOT 3330)
Department
of Computer Science
Seminar Series
Doctor AI: Interpretable Deep Learning Models for Healthcare Applications
Dr. Jimeng Sun
Associate Professor of Computational Science & Engineering
Georgia Institute of Technology
Thursday, September 29, 2016 11:15am to 12:15pm
Venue: ISC Room 3300
Abstract: When you face medical problems, do you want to be seen by a person or a machine? We will present our efforts in making such a machine that can learn from human doctors based on their collective memory, i.e., large amount of Electronic Health Record (EHR) data. We cover three aspects:
Temporality: We develop an application of Recurrent Neural Networks (RNN) to jointly forecast the future disease diagnoses. Our model assesses the entire history of patients and make continuous and multilabel predictions based on patients' historical data [1].
Representation: We propose Med2Vec, two-level distributional representation for medical codes including diagnoses, medications and procedures [2].
Interpretation: Most existing works suffer the tradeoff between interpretation and accuracy by either picking complex black box models such as RNN or relying on less accurate traditional models with better interpretation such as logistic regression. To address this dilemma, we present REverse Time AttentIoN model (RETAIN) for analyzing Electronic Health Records (EHR) data that achieves high accuracy while remaining clinically interpretable [3].
Bio:
Jimeng Sun is an Associate Professor at College of Computing in Georgia Institute of Technology. Prior to joining Georgia Tech, he was a research staff member at IBM TJ Watson
Research Center. His research focuses on health analytics using electronic health records and data mining, especially in designing deep learning models, tensor analysis and similarity learning methods and large-scale predictive modeling systems. He has published
over 100 papers, filed over 20 patents (5 granted). He has received ICDM best research paper award in 2008, SDM best research paper award in 2007, and KDD Dissertation runner-up award in 2008. Dr. Sun received his B.S. and M.Phil. in Computer Science from
Hong Kong University of Science and Technology in 2002 and 2003, and PhD in Computer Science from Carnegie Mellon University in 2007.
[1] Edward Choi, Mohammad Bahadori, Elizabeth Searles, Catherine Coffey, Michael Thompson, James Bost, Javier Tejedor-Sojo, and Jimeng Sun. “Multi-layer Representation Learning for Medical Concepts.” KDD 16.
[2] Edward Choi, Mohammad Taha Bahadori, Andy Schuetz, Walter F. Stewart, and Jimeng Sun, "Doctor AI: Predicting Clinical Events via Recurrent Neural Networks", arXiv:1511.05942 [cs.LG]
[3] Edward Choi, Mohammad Taha Bahadori, Andy Schuetz, Walter F. Stewart, Jimeng Sun. "RETAIN: Interpretable Predictive Model in Healthcare using Reverse Time Attention Mechanism", NIPS'16
(Refreshments will be served)
