Umar Syed

Research Scientist

Google

111 8th Avenue

New York, NY 10011

I am a research scientist at Google. I received a Ph.D. in Computer Science from Princeton University, where I was advised by Rob Schapire. I spent two years as a postdoctoral researcher at the University of Pennsylvania, hosted by Ben Taskar and Michael Kearns.

Curriculum Vitae

Publications

• • Berke, A., Ghazi, B., Bacis, E., Kamath, P., Kumar, R., Lassonde, R., Manurangsi, P., Syed, U. (2025) How unique is whose web browser? The role of demographics in browser fingerprinting among US users. Proceedings on Privacy Enhancing Technologies (PoPETs 2025).

  • Amin, K., Bie, A., Kong, W., Kurakin, A., Ponomareva, N., Syed, U., Terzis, A., Vassilvitskii, S. (2024)* Private prediction for large-scale synthetic text generation. Findings of the Association for Computational Linguistics (EMNLP 2024).

  • Kong, W., Medina, A. M., Ribero, M., Syed, U. (2024)* DP-Auditorium: A large-scale library for auditing differential privacy. Proceedings of the IEEE Symposium on Security and Privacy (S&P 2024).

  • Bacis, E., Bilogrevic, I., Busa-Fekete, R., Herath, A., Sartori, A., Syed, U. (2024)* Assessing web fingerprinting risk. Companion Proceedings of the ACM Web Conference (WWW 2024). 

  • Abernethy, J., Schapire, R. E., Syed, U. (2024)* Lexicographic optimization: Algorithms and stability. Proceedings of the Twenty-Seventh International Conference on Artificial Intelligence and Statistics (AISTATS 2024). 

  • Busa-Fekete, R., Medina, A. M., Syed, U., Vassilvitskii, S. (2023)* Label differential privacy and private training data release. Proceedings of the Fortieth International Conference on Machine Learning (ICML 2023).

  • Bravo-Hermsdorff, G., Busa-Fekete, R., Ghavamzadeh, M., Medina, A. M., Syed, U. (2022)* Private and communication-efficient algorithms for entropy estimation. Advances in Neural Information Processing Systems 35 (NeurIPS 2022).

  • Esfandiari, H., Mirrokni, V., Syed, U., Vassilvitskii, S. (2021)* Label differential privacy via clustering. Proceedings of the Twenty-Fifth International Conference on Artificial Intelligence and Statistics (AISTATS 2022).

  • Epasto, A., Medina, A. M.,  Avery, S., Bai, Y., Busa-Fekete, R., Carey, C., Gao, Y., Guthrie, D., Ghosh, S., Ioannidis, J., Jiao, J., Lacki, J., Lee, J., Mauser, A., Milch, B., Mirrokni, V., Ravichandran, D., Shi, W., Spero, M., Sun, Y., Syed, U., Vassilvtiskii, S., Wang, S. (2021) Clustering for private interest-based advertising. Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining (KDD 2021).

  • Medina, A. M.,  Syed, U., Vassilvitskii, S., Vitercik, E. (2021)* Private optimization without constraint violations. Proceedings of the Twenty-Fourth International Conference on Artificial Intelligence and Statistics (AISTATS 2021).

  • Syed, U., Vassilvitskii, S. (2017)* SQML: Large-scale in-database machine learning with pure SQL. Proceedings of the 2017 Symposium on Cloud Computing (SoCC 2017).

  • Balkanski, E., Syed, U., Vassilvitskii, S. (2017)* Statistical cost sharing. Advances in Neural Information Processing Systems 30 (NeurIPS 2017).

  • Nazerzadeh, H., Paes Leme, R., Rostamizadeh, A., Syed, U. (2016)* Where to sell: Simulating auctions from learning algorithms. Proceedings of the Seventeenth ACM Conference on Economics and Computation (EC 2016).

  • Heidari, H., Mahdian, M., Syed, U., Vassilvitskii, S., Yazdanbod, S. (2016)* Pricing a low-regret seller. Proceedings of the Thirty-Third International Conference on Machine Learning (ICML 2016).

  • Choromanski, K., Rostamizadeh, A., Syed, U. (2015)* An optimal online algorithm for retrieving heavily perturbed statistical databases in the low-dimensional querying model. Proceedings of the Twenty-Fourth ACM International Conference on Information and Knowledge Management (CIKM 2015).

  • DeSalvo, G., Mohri, M., Syed, U. (2015)* Learning with deep cascades. Proceedings of the Twenty-Sixth International Conference on Algorithmic Learning Theory (ALT 2015).

  • Cortes, C., Kuznetsov, V., Mohri, M., Syed, U. (2015)* Structural maxent models. Proceedings of the Thirty-Second International Conference on Machine Learning (ICML 2015). Supplement.

  • Kuznetsov, V., Mohri, M., Syed, U. (2014)* Multi-class deep boosting. Advances in Neural Information Processing Systems 27 (NeurIPS 2014). Supplement.

  • Amin, K., Rostamizadeh, A., Syed, U. (2014)* Repeated contextual auctions with strategic buyers. Advances in Neural Information Processing Systems 27 (NeurIPS 2014). Supplement.

  • Cortes, C., Mohri, M., Syed, U. (2014)* Deep boosting. Proceedings of the Thirty-First International Conference on Machine Learning (ICML 2014). Supplement.

  • Asuncion, A., de Haan, J., Mohri, M., Patel, K., Rostamizadeh, A., Syed, U., Wong, L. (2014)* Corporate learning at scale: Lessons from a large online course at Google. Learning at Scale 2014.

  • Amin, K., Rostamizadeh, A., Syed, U. (2013)* Learning prices for repeated auctions with strategic buyers. Advances in Neural Information Processing Systems 26 (NeurIPS 2013). Supplement.

  • Amin, K., Kearns, M., Syed, U. (2011)* Graphical models for bandit problems. Uncertainty in Artificial Intelligence: Proceedings of the Twenty-Seventh Conference (UAI 2011).

  • Amin, K., Kearns, M., Syed, U. (2011)* Bandits, query learning, and the haystack dimension. Proceedings of the Twenty-Fourth Annual Conference on Learning Theory (COLT 2011).

  • Fabrikant, A., Syed, U., Rexford, J. (2011) There's something about MRAI: Timing diversity can exponentially worsen BGP convergence. Proceedings of the Thirtieth IEEE International Conference on Computer Communications (INFOCOM 2011).

  • Syed, U., Taskar, B. (2010) Semi-supervised learning with adversarially missing label information. Advances in Neural Information Processing Systems 23 (NeurIPS 2010). Supplement.

  • Syed, U., Schapire, R. E. (2010) A reduction from apprenticeship learning to classification. Advances in Neural Information Processing Systems 23 (NeurIPS 2010).

  • Brautbar, M., Kearns, M., Syed, U. (2010)* Private and third-party randomization in risk-sensitive equilibrium concepts. Proceedings of the Twenty-Fourth Conference on Artificial Intelligence (AAAI 2010).

  • Syed, U. (2010) Reinforcement learning without rewards. Ph.D. dissertation. Department of Computer Science, Princeton University.

  • Syed, U., Slivkins, A., Mishra, N. (2009) Adapting to the shifting intent of search queries. Advances in Neural Information Processing Systems 22 (NeurIPS 2009). Full version

  • Syed, U., Bowling, M., Schapire, R. E. (2008) Apprenticeship learning using linear programming. Proceedings of the Twenty-Fifth International Conference on Machine Learning (ICML 2008).

  • Syed, U., Williams, J. D. (2008) Using automatically transcribed dialogs to learn user models in a spoken dialog system. Proceedings of the Forty-Sixth Annual Meeting of the Association for Computational Linguistics (ACL 2008).

  • Syed, U., Yona, G. (2008) Enzyme function prediction with interpretable models. Methods in Molecular Biology: Computational Systems Biology, edited by J. McDermott, K. Montgomery, R. Bumgarner and R. Samudrala. Humana Press. [Book chapter]

  • Syed, U., Schapire, R. E. (2007) A game-theoretic approach to apprenticeship learning. Advances in Neural Information Processing Systems 20 (NeurIPS 2007). Supplement and videos.

  • Syed, U., Schapire, R. E. (2007) Imitation learning with a value-based prior. Uncertainty in Artificial Intelligence: Proceedings of the Twenty-Third Conference (UAI 2007).

  • Syed, U., Yona, G. (2003) Using a mixture of probabilistic decisions trees for direct prediction of protein function. Proceedings of the Seventh Annual International Conference on Research in Computational Biology (RECOMB 2003).

* Denotes alphabetical author order.

Code

• Apprenticeship learning code from NeurIPS 2007 and ICML 2008 papers (includes highway simulator).

• Deep boosting code from ICML 2014 paper.