David Sontag's Home Page
E-mail: dsontag {@ | at} mit.edu
Clinical machine learning group website
I am a Professor of
Electrical Engineering and
Computer Science at MIT, part of
the Institute for Medical
Engineering & Science,
the Computer Science and Artificial
Intelligence Laboratory, and the J-Clinic for Machine
Learning in Health.
My research focuses on advancing machine learning and artificial intelligence, and using these to transform health care.
Previously, I was an Assistant Professor of Computer Science and
Data Science at New York University.
News
- I will be on leave from MIT for all of 2024 and am CEO of
Layer Health, which I co-founded
with several former MIT students.
- Our MIT Machine Learning for Healthcare class is available
on MIT
OpenCourseWare
(all videos).
Teaching
Spring '17, '19, '20, '21, '22: Machine Learning
for Healthcare (6.7930, HST.956)
Fall '20, '21, '22: Introduction to Machine Learning (6.036)
Fall '17, '18, '19: Machine Learning (6.867)
Fall 2016: Inference and Representation (DS-GA-1005 and CSCI-GA.2569)
Spring 2016: Introduction to Machine Learning (CSCI-UA.0480-007)
Selected papers:
- S. Hegselmann, A. Buendia, H. Lang, M. Agrawal, X. Jiang,
D. Sontag. TabLLM: Few-shot Classification of Tabular Data with Large Language
Models. 26th International Conference on Artificial
Intelligence and Statistics (AISTATS), 2023.
- M. Agrawal, S. Hegselmann, H. Lang, Y. Kim,
D. Sontag. Large
Language Models are Few-Shot Clinical Information
Extractors. Conference on Empirical Methods in Natural
Language Processing (EMNLP), 2022.
- H. Lang, M. Agrawal, Y. Kim,
D. Sontag. Co-training
Improves Prompt-based Learning for Large Language
Models. ICML, 2022.
- H. Mozannar, A. Satyanarayan,
D. Sontag. Teaching
Humans When To Defer to a Classifier via
Exemplars. AAAI, 2022.
- L. Murray, D. Gopinath, M. Agrawal, S. Horng, D.
Sontag,
D. Karger. MedKnowts:
Unified Documentation and Information Retrieval for Electronic
Health Records. UIST, 2021. (Video)
- Z. Hussain, R. Krishnan,
D. Sontag. Neural
Pharmacodynamic State Space Modeling. ICML, 2021.
- R. Kodialam, R. Boiarsky, J. Lim, N. Dixit, A. Sai,
D. Sontag. Deep
Contextual Clinical Prediction with Reverse
Distillation. AAAI, 2021.
- M. Oberst, D. Sontag. Counterfactual Off-Policy Evaluation with
Gumbel-Max Structural Causal Models, ICML 2019.
- I. Chen, F. Johansson,
D. Sontag. Why Is My
Classifier Discriminatory?, NeurIPS, 2018.
- U. Shalit, F. Johansson, D. Sontag. Estimating Individual Treatment Effect: Generalization Bounds and Algorithms. 34th International Conference on Machine Learning (ICML), 2017. [code] [Slides]
- M. Rotmensch, Y. Halpern, A. Tlimat, S. Horng,
D. Sontag. Learning a Health Knowledge Graph from Electronic Medical Records, Nature Scientific Reports, July 2017. Supplementary
- R. Krishnan, U. Shalit, D. Sontag. Structured Inference Networks for Nonlinear State Space Models, Thirty-First AAAI Conference on Artificial Intelligence, Feb. 2017. [code] Older version
- Y. Halpern, S. Horng, Y. Choi,
D. Sontag. Electronic
Medical Record Phenotyping using the Anchor and Learn
Framework. Journal of the American Medical Informatics
Association (JAMIA), April
2016. [html]
[code] [Slides]
- Y. Kim, Y. Jernite, D. Sontag, S. Rush. Character-Aware Neural Language Models, Thirtieth AAAI Conference on Artificial Intelligence, Feb. 2016. [code] [Slides] Video
- X. Wang, D. Sontag, F. Wang. Unsupervised Learning of Disease Progression Models. ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD), Aug. 2014. [Slides] BibTex
- E. Brenner, D. Sontag. SparsityBoost: A New Scoring Function for Learning Bayesian Network Structure. Uncertainty in Artificial Intelligence (UAI) 29, July 2013. BibTex [arXiv]
- S. Arora, R. Ge, Y. Halpern, D. Mimno, A. Moitra, D. Sontag, Y. Wu, M. Zhu. A Practical Algorithm for Topic Modeling with Provable Guarantees. 30th International Conference on Machine Learning (ICML), 2013. Supplementary BibTex
- T. Koo, A. Rush, M. Collins, T. Jaakkola, and D. Sontag. Dual Decomposition for Parsing with Non-Projective Head Automata. Empirical Methods in Natural Language Processing (EMNLP), 2010. Best paper award. BibTex
- T. Jaakkola, D. Sontag, A. Globerson,
M. Meila. Learning
Bayesian Network Structure using LP Relaxations. 13th International Conference on Artificial Intelligence
and Statistics (AI-STATS),
2010. BibTex
- D. Sontag. Approximate
Inference in Graphical Models using LP
Relaxations. Ph.D. thesis, Massachusetts Institute of Technology, 2010.
George M. Sprowls Award for the best doctoral theses in Computer
Science at MIT (2010). BibTex
- D. Sontag, T. Meltzer, A. Globerson, Y. Weiss, T. Jaakkola. Tightening
LP Relaxations for MAP using Message Passing. Uncertainty
in Artificial Intelligence (UAI) 24, July 2008. Best paper award. [code] BibTex
- D. Sontag, T. Jaakkola. New
Outer Bounds on the Marginal Polytope. Neural Information Processing Systems
(NIPS) 20, Dec. 2007. Outstanding student paper award. Addendum BibTex
Code (for latest, see
our Github repo)
Download Python code for learning topic models (corresponds to ICML '13 paper). See also David Mimno's Mallet-compatible Java implementation.
Download code for learning Bayesian network structure (corresponds to UAI '13 SparsityBoost paper).
Download C++ code for MAP inference in graphical models (corresponds to
UAI '12 paper; see readme file).
Low-dimensional embeddings of
medical concepts (corresponds to AMIA CRI '16 paper)
DeepDiagnosis from longitudinal clinical data (corresponds to MLHC '16 paper)
omop-learn, Python package
for deep learning on longitudinal health data