Anand D. Sarwate

CoRE Building Rm. 517

Busch Campus

Rutgers University

Piscataway, NJ 08854-8058

I am an Associate Professor in the Department of Electrical and Computer Engineering at Rutgers, The State University of New Jersey. I also am a member of the graduate faculty in the Department of Computer Science and the Department of Statistics. I like to work on problems that involve probability, mathematical statistics,
and optimization, with applications in information theory, communication, signal processing, and machine learning. I’m particularly interested in how these things intersect in the context of distributed/decentralized systems with constraints like privacy, bandwidth, latency, power, and so on.

This page is still somewhat under construction. I anticipate that will be the state for quite a while.

news

May 28, 2023	Just returned from a really fabulous workshop on Information-Theoretic Methods for Trustworthy Machine Learning at the Simons Institute for the Theory of Computing at UC Berkeley. It was weird to wander around a place I lived 20 years ago but really nice to see old friends and meet new people.
May 19, 2023	Finally mostly ported everything over to the new site…
Jul 1, 2022	New grant! RINGS: REALTIME: Resilient Edge-cloud Autonomous Learning with Timely Inferences, co-PIs Roy D. Yates, Waheed U. Bajwa, and Dipankar Raychaudhuri. We will be looking at issues in adaptive edge-cloud assisted ML-based services from the ground-up.

selected publications

R. Islam, K. N. Keya, S. Pan, A. D. Sarwate, and J. R. Foulds, “Differential Fairness: An Intersectional Framework for Fair AI,” Entropy, vol. 25, no. 4, Apr. 2023.
Abs Bib DOI

We propose definitions of fairness in machine learning and artificial intelligence systems that are informed by the framework of intersectionality, a critical lens from the legal, social science, and humanities literature which analyzes how interlocking systems of power and oppression affect individuals along overlapping dimensions including gender, race, sexual orientation, class, and disability. We show that our criteria behave sensibly for any subset of the set of protected attributes, and we prove economic, privacy, and generalization guarantees. Our theoretical results show that our criteria meaningfully operationalize AI fairness in terms of real-world harms, making the measurements interpretable in a manner analogous to differential privacy. We provide a simple learning algorithm using deterministic gradient methods, which respects our intersectional fairness criteria. The measurement of fairness becomes statistically challenging in the minibatch setting due to data sparsity, which increases rapidly in the number of protected attributes and in the values per protected attribute. To address this, we further develop a practical learning algorithm using stochastic gradient methods which incorporates stochastic estimation of the intersectional fairness criteria on minibatches to scale up to big data. Case studies on census data, the COMPAS criminal recidivism dataset, the HHP hospitalization data, and a loan application dataset from HMDA demonstrate the utility of our methods.
```
@article{IslamKPSF:23df,
  article-number = {660},
  author = {Islam, Rashidul and Keya, Kamrun Naher and Pan, Shimei and Sarwate, Anand D. and Foulds, James R.},
  doi = {10.3390/e25040660},
  issn = {1099-4300},
  journal = {Entropy},
  month = apr,
  number = {4},
  title = {Differential Fairness: An Intersectional Framework for Fair AI},
  url = {https://www.mdpi.com/1099-4300/25/4/660},
  volume = {25},
  year = {2023},
}
```

Z. Wang, A. Engel, A. Sarwate, I. Dumitriu, and T. Chiang, “Spectral evolution and invariance in linear-width neural networks,” ArXiV, arXiv:2211.06506 [cs.LG], Nov. 2022.

Bib DOI

@techreport{WangWSDC:22spectral,
  author = {Wang, Zhichao and Engel, Andrew and Sarwate, Anand and Dumitriu, Ioana and Chiang, Tony},
  doi = {10.48550/arXiv.2211.06506},
  institution = {ArXiV},
  month = nov,
  number = {arXiv:2211.06506 [cs.LG]},
  title = {Spectral evolution and invariance in linear-width neural networks},
  year = {2022},
}

H. Gazula, K. Rootes-Murdy, B. Holla, S. Basodi, Z. Zhang, E. Verner, R. Kelly, P. Murthy, A. Chakrabarti, D. Basu, S. Bhagyalakshmi Nanjayya, R. Lenin Singh, R. Lourembam Singh, K. Kalyanram, K. Kartik, K. Kalyanaraman, K. Ghattu, R. Kuriyan, S. S. Kurpad, G. J. Barker, R. D. Bharath, S. Desrivieres, M. Purushottam, D. P. Orfanos, E. Sharma, M. Hickman, M. Toledano, N. Vaidya, T. Banaschewski, A. L. W. Bokde, H. Flor, A. Grigis, H. Garavan, P. Gowland, A. Heinz, R. Brühl, J.-L. Martinot, M.-L. Paillére Martinot, E. Artiges, F. Nees, T. Paus, L. Poustka, J. H. Fröhner, L. Robinson, M. N. Smolka, H. Walter, J. Winterer, R. Whelan, J. A. Turner, A. D. Sarwate, et al., “Federated Analysis in COINSTAC Reveals Functional Network Connectivity and Spectral Links to Smoking and Alcohol Consumption in Nearly 2,000 Adolescent Brains,” Neuroinformatics, vol. 21, pp. 287–301, 2022.
Abs Bib DOI

With the growth of decentralized/federated analysis approaches in neuroimaging, the opportunities to study brain disorders using data from multiple sites has grown multi-fold. One such initiative is the Neuromark, a fully automated spatially constrained independent component analysis (ICA) that is used to link brain network abnormalities among different datasets, studies, and disorders while leveraging subject-specific networks. In this study, we implement the neuromark pipeline in COINSTAC, an open-source neuroimaging framework for collaborative/decentralized analysis. Decentralized exploratory analysis of nearly 2000 resting-state functional magnetic resonance imaging datasets collected at different sites across two cohorts and co-located in different countries was performed to study the resting brain functional network connectivity changes in adolescents who smoke and consume alcohol. Results showed hypoconnectivity across the majority of networks including sensory, default mode, and subcortical domains, more for alcohol than smoking, and decreased low frequency power. These findings suggest that global reduced synchronization is associated with both tobacco and alcohol use. This proof-of-concept work demonstrates the utility and incentives associated with large-scale decentralized collaborations spanning multiple sites.
```
@article{Gazula:22federated,
  author = {Gazula, Harshvardhan and Rootes-Murdy, Kelly and Holla, Bharath and Basodi, Sunitha and Zhang, Zuo and Verner, Eric and Kelly, Ross and Murthy, Pratima and Chakrabarti, Amit and Basu, Debasish and Bhagyalakshmi Nanjayya, Subodh and Lenin Singh, Rajkumar and Lourembam Singh, Roshan and Kalyanram, Kartik and Kartik, Kamakshi and Kalyanaraman, Kumaran and Ghattu, Krishnaveni and Kuriyan, Rebecca and Kurpad, Sunita Simon and Barker, Gareth J. and Bharath, Rose Dawn and Desrivieres, Sylvane and Purushottam, Meera and Orfanos, Dimitri Papadopoulos and Sharma, Eesha and Hickman, Matthew and Toledano, Mireille and Vaidya, Nilakshi and Banaschewski, Tobias and Bokde, Arun L. W. and Flor, Herta and Grigis, Antoine and Garavan, Hugh and Gowland, Penny and Heinz, Andreas and Br{\"u}hl, R{\"u}diger and Martinot, Jean-Luc and Paill{\'e}re Martinot, Marie-Laure and Artiges, Eric and Nees, Frauke and Paus, Tom{\'a}s and Poustka, Luise and Fr{\"o}hner, Juliane H. and Robinson, Lauren and Smolka, Michael N. and Walter, Henrik and Winterer, Jeanne and Whelan, Robert and Turner, Jessica A. and Sarwate, Anand D. and Plis, Sergey M. and Benegal, Vivek and Schumann, Gunter and Calhoun, Vince D. and {IMAGEN Consortium}},
  date = {2022/11/25},
  doi = {10.1007/s12021-022-09604-4},
  id = {Gazula2022},
  isbn = {1559-0089},
  journal = {Neuroinformatics},
  month = apr,
  pages = {287--301},
  title = {Federated Analysis in COINSTAC Reveals Functional Network Connectivity and Spectral Links to Smoking and Alcohol Consumption in Nearly 2,000 Adolescent Brains},
  url = {https://doi.org/10.1007/s12021-022-09604-4},
  volume = {21},
  year = {2023},
}
```

Y. Zhang, S. Jaggi, M. Langberg, and A. D. Sarwate, “The Capacity of Causal Adversarial Channels,” in Proceedings of the 2022 IEEE International Symposium on Information Theory (ISIT), 2022.

arXiv Bib DOI

@inproceedings{ZhangJLS:22isit,
  author = {Zhang, Yihan and Jaggi, Sidharth and Langberg, Michael and Sarwate, Anand D.},
  booktitle = {Proceedings of the 2022 IEEE International Symposium on Information Theory (ISIT)},
  doi = {10.1109/ISIT50566.2022.9834709},
  month = jun,
  title = {The Capacity of Causal Adversarial Channels},
  url = {https://doi.org/10.1109/ISIT50566.2022.9834709},
  year = {2022},
}

N. Sathyavageeswaran, R. D. Yates, A. D. Sarwate, and N. Mandayam, “Privacy Leakage in Discrete Time Updating Systems,” in Proceedings of the 2022 IEEE International Symposium on Information Theory (ISIT), 2022.

arXiv Bib DOI

@inproceedings{SathyavageeswaranYSM:22isit,
  author = {Sathyavageeswaran, Nitya and Yates, Roy D. and Sarwate, Anand D. and Mandayam, Narayan},
  booktitle = {Proceedings of the 2022 IEEE International Symposium on Information Theory (ISIT)},
  doi = {10.1109/ISIT50566.2022.9834673},
  month = jun,
  title = {Privacy Leakage in Discrete Time Updating Systems},
  url = {https://doi.org/10.1109/ISIT50566.2022.9834673},
  year = {2022},
}

Y. Zhang, S. Vatedka, S. Jaggi, and A. D. Sarwate, “Quadratically Constrained Myopic Adversarial Channels,” IEEE Transactions on Information Theory, vol. 68, pp. 4901–4948, Aug. 2022.

Bib DOI

@article{ZhangVJS:22myopicgavc,
  author = {Zhang, Yihan and Vatedka, Shashank and Jaggi, Sidharth and Sarwate, Anand D.},
  doi = {10.1109/TIT.2022.3167554},
  journal = {IEEE Transactions on Information Theory},
  month = aug,
  pages = {4901--4948},
  title = {Quadratically Constrained Myopic Adversarial Channels},
  url = {https://doi.org/10.1109/TIT.2022.3167554},
  volume = {68},
  year = {2022},
}

S. Li, P. Krishan, S. Jaggi, M. Langberg, and A. D. Sarwate, “Computationally Efficient Codes for Adversarial Binary-Erasure Channels,” in Proceedings of the 2023 IEEE International Symposium on Information Theory (ISIT), 2023, pp. to appear.

Bib

@inproceedings{LKJLS:23efficient,
  author = {Li, Sijie and Krishan, Prasad and Jaggi, Sidharth and Langberg, Michael and Sarwate, Anand D},
  booktitle = {Proceedings of the 2023 IEEE International Symposium on Information Theory (ISIT)},
  month = jun,
  pages = {to appear},
  title = {Computationally Efficient Codes for Adversarial Binary-Erasure Channels},
  year = {2023},
}