- Ph.D., Electrical and Computer Engineering, Carnegie Mellon University
- M.S., Electrical and Computer Engineering, Carnegie Mellon University
- B. Tech., Electronics and Electrical Communication Engineering, Indian Institute of Technology, Kharagpur
- 2022 Simons Institute Fellowship for Causality
- 2021 A G Milnes Outstanding Thesis Award
- 2020 Cylab Presidential Fellowship
- 2019 K&L Gates Presidential Fellowship in Ethics and Computational Technologies
- 2019 Axel Berny Presidential Graduate Fellowship
- 2017 Tan Endowed Graduate Fellowship
- 2016 Prabhu and Poonam Goel Graduate Fellowship
- 2015 Nilanjan Ganguly Memorial Award for Best B. Tech. Thesis
- 2014 HONDA Young Engineer and Scientist Award
Dr. Sanghamitra Dutta joined the University of Maryland, College Park in 2022. Her research interests broadly revolve around reliable and trustworthy machine learning. She is particularly interested in addressing the challenges concerning fairness, explainability, and law, by bringing in a novel foundational perspective deep-rooted in information theory, causality, and optimization.
Prior to joining UMD, she was a senior research associate at JPMorgan Chase AI Research in the Explainable AI Centre of Excellence (XAI CoE). Her research on algorithmic fairness has been adopted as part of the fair lending model review at JPMorgan.
Dr. Dutta received her Ph. D. from Carnegie Mellon University. Her thesis proposes a systematic quantification of the legally non-exempt disparity in machine learning models, bringing together causality, information theory, and law. She has received the 2019 K&L Gates Presidential Fellowship in Ethics and Computational Technologies for her research in this direction. Her research on quantifying accuracy-fairness tradeoffs using information theory (with IBM Research) was featured in New Scientist. Her thesis received the 2021 A G Milnes Outstanding Thesis Award.
In her prior work, she has also examined problems in reliable computing, proposing novel algorithmic solutions for large-scale distributed machine learning, using tools from coding theory (an emerging area called “coded computing”). Her results on coded computing address problems in algorithm-based-fault-tolerance that have been open for several decades and have received substantial attention from across disciplines. She has also pursued research internships at IBM Research and Dataminr.
Her research vision is to build the foundations of reliable artificial intelligence (AI), beginning from a fundamental understanding of the challenges in reliability and trust, and carrying them all the way to practical implementations, so that AI can truly bring about social good.
Her research interests include:
- Fairness and Explainability
- Information Theory
- Coded Computing
ENEE436: Foundations of Machine Learning Fall 2022
- S. Dutta, J Long, S Mishra, C Tilli, D Magazzeni, "Robust Counterfactual Explanations for Tree-Based Ensembles," International Conference on Machine Learning (ICML 2022).
- P. Mathur, A T Neerkaje, M Chhibber, R Sawhney, F Guo, F Dernoncourt, S Dutta, D Manocha, "MONOPOLY: Financial Prediction from MONetary POLicY Conference Videos Using Multimodal Cues," ACM Multimedia 2022 (ACM-MM 2022).
- S Dutta, P Venkatesh, P Grover, "Quantifying Feature Contributions to Overall Disparity Using Information Theory," AAAI-22 Workshop on Information-Theoretic Methods for Causal Inference and Discovery (AAAI Workshop 2022).
- P Venkatesh, S Dutta*, N Mehta*, P Grover, "Can Information Flows Suggest Targets for Interventions in Neural Circuits?," Neural Information Processing Systems (NeurIPS 2021).
- S. Dutta, P. Venkatesh, P. Mardziel, A. Datta and P. Grover, "Fairness under Feature Exemptions: Counterfactual and Observational Measures," IEEE Transactions on Information Theory 2021.
- S. Dutta, J. Wang, and G. Joshi, "Slow and stale gradients can win the race," IEEE Journal on Selected Areas in Information Theory 2021.
- S Mishra, S Dutta, J Long, D Magazzeni, "A Survey on the Robustness of Feature Importance and Counterfactual Explanations," Explainable AI in Finance (XAI-FIN21).
- C. Jiang*, B. Wu*, S. Dutta and P. Grover, "Bursting the Bubbles: Debiasing Recommendation Systems While Allowing for Chosen Category Exemptions," BIAS Workshop at ECIR (ECIR Workshop 2021).
- S. Dutta, L. Ma, T. K. Saha, D. Liu, J. Tetreault, and A. Jaimes, "GTN-ED: Event Detection Using Graph Transformer Networks," TextGraphs Workshop at NAACL (NAACL Workshop 2021).
- S. Dutta, D. Wei, H. Yueksel, P. Y. Chen, S. Liu, and K. R. Varshney, "Is There a Trade-Off Between Fairness and Accuracy? A Perspective Using Mismatched Hypothesis Testing," International Conference on Machine Learning (ICML 2020).
- S. Dutta, P. Venkatesh, P. Mardziel, A. Datta and P. Grover, "An Information-Theoretic Quantification of Discrimination with Exempt Features," AAAI Conference on Artificial Intelligence (AAAI 2020, ORAL).
- P. Venkatesh, S. Dutta, and P. Grover, "How else should we define Information Flow in Neural Circuits,” IEEE International Symposium on Information Theory (ISIT 2020).
- P. Venkatesh, S. Dutta, and P. Grover, "Information Flow in Computational Systems,” IEEE Transactions on Information Theory, Sep 2020.
- S. Dutta*, M. Fahim*, H. Jeong*, F. Haddadpour*, V. Cadambe, and P. Grover, "On the Optimal Recovery Threshold of Coded Matrix Multiplication," IEEE Transactions on Information Theory, Jan 2020.
- S. Dutta*, H. Jeong*, Y. Yang*, V. Cadambe, T. M. Low and P. Grover, “Addressing Unreliability in Emerging Devices and Non-von Neumann Architectures Using Coded Computing," Proceedings of the IEEE, April 2020.
- P. Venkatesh, S. Dutta and P. Grover, "How should we define Information Flow in Neural Circuits,” IEEE International Symposium on Information Theory (ISIT 2019).
- S. Dutta, V. Cadambe and P. Grover, "Short-Dot: Computing Large Linear Transforms Distributedly using Coded Short Dot Products," IEEE Transactions on Information Theory, Oct 2019.
- S. Dutta, Z. Bai, T. M. Low and P. Grover, "CodeNet: Training Large Scale Neural Networks in Presence of Soft-Errors," Coding Theory For Large-scale Machine Learning Workshop at ICML (CodML Workshop, ICML 2019, Spotlight).
- S. Dutta, G. Joshi, P. Dube, S. Ghosh, and P. Nagpurkar, "Slow and stale gradients can win the race: Error-Runtime trade-offs in Distributed SGD," International Conference on Artificial Intelligence and Statistics (AISTATS 2018).
- U. Sheth, S. Dutta, M. Chaudhari, H. Jeong, Y. Yang, J. Kohonen, T. Roos, and P. Grover, "An Application of Storage-Optimal MatDot Codes for Coded Matrix Multiplication: Fast k-Nearest Neighbors Estimation,” IEEE International Conference on Big Data (IEEE BigData 2018).
- S. Dutta*, Z. Bai*, H. Jeong, T. M. Low, and P. Grover, "A Unified Coded Deep Neural Network Training Strategy based on Generalized PolyDot Codes," IEEE International Symposium on Information Theory (ISIT 2018).
- S. Dutta, V. Cadambe and P. Grover, "Coded Convolution for parallel and distributed computing within a deadline," IEEE International Symposium on Information Theory (ISIT 2017).
- M. Fahim*, H. Jeong*, F. Haddadpour, S. Dutta, V. Cadambe, and P. Grover, "On the Optimal Recovery Threshold of Coded Matrix Multiplication," Communication, Control and Computing (Allerton 2017).
- S. Dutta, V. Cadambe and P. Grover, "Short-Dot: Computing Large Linear Transforms Distributedly using Coded Short Dot Products," Neural Information Processing Systems (NeurIPS 2016).
- S. Dutta and P. Grover, "Adaptivity provably helps: Information-theoretic limits on l0 cost of non-adaptive sensing," IEEE International Symposium on Information Theory (ISIT 2016).
- S. Dutta, Y. Yang, N. Wang, E. Pop, V. Cadambe and P. Grover, “Reliable Matrix Multiplication using Error-prone Dot-product Nanofunctions with an application to logistic regression” (SRC Techcon, 2016).
- S. Dutta and A. De, "Sparse UltraWideBand Radar Imaging in a Locally Adapting Matching Pursuit (LAMP) Framework," IEEE International Radar Conference (RADAR 2015).