Omid Sadeghi

I'm a final-year Ph.D. candidate in the Electrical and Computer Engineering department at the University of Washington, supervised by Prof. Maryam Fazel. I've also recently obtained my Master's degree in Mathematics (with a focus on mathematical optimization) here at UW. Before coming to UW, I did my bachelor's at the Sharif University of Technology, where I obtained B.S. degrees in Electrical Engineering and Mathematics. Also, during my undergraduate studies, I spent the summer 2015 as a junior research assistant at the Chinese University of Hong Kong (CUHK) under the supervision of Prof. Chandra Nair.

I specialize in applying convex optimization tools to tackle non-convex problems (particularly continuous submodular maximization) in Machine Learning under various additional considerations (e.g., limited resource/budget availability, privacy, incentive compatibility, and fairness) and in both online and offline settings. My research finds applications in online advertising and online resource allocation problems.

Competencies: Python, CVXPY, Scikit-learn, Pandas.

In my spare time, you can find me learning new languages (currently learning German and Spanish), going for a run, or playing soccer. Also, I am a certified Heroic coach.

Research Interests

• Online Optimization and Learning

• Submodular Optimization

• Privacy and Incentives in Machine Learning

Publications (Google Scholar)

• No-Regret Online Prediction with Strategic Experts,
  With Maryam Fazel, Accepted for NeurIPS 2023.

• Online SuBmodular + SuPermodular (BP) Maximization with Bandit Feedback,
  With Adhyyan Narang, Lillian J Ratliff, Maryam Fazel, and Jeff Bilmes, Under Review.

• Function Design for Improved Competitive Ratio in Online Resource Allocation with Procurement Costs,
  With Mitas Ray, Lillian J. Ratliff, and Maryam Fazel, Under Review.

• Fast First-Order Methods for Monotone Strongly DR-Submodular Maximization,
  With Maryam Fazel, SIAM Conference on Applied and Computational Discrete Algorithms (ACDA23) .

• Improved Regret Bounds for Online Submodular Maximization,
  With Prasanna Raut and Maryam Fazel, ICML 2021 Workshop on Subset Selection in Machine Learning: From Theory to Applications.

• Differentially Private Monotone Submodular Maximization Under Matroid and Knapsack Constraints,
  With Maryam Fazel, AISTATS 2021. [Talk], [Poster]

• Online DR-Submodular Maximization: Minimizing Regret and Constraint Violation,
  With Prasanna Raut and Maryam Fazel, AAAI 2021. [Talk], [Poster], [Full Version]

• A Single Recipe for Online Submodular Maximization with Adversarial or Stochastic Constraints,
  With Prasanna Raut and Maryam Fazel, NeurIPS 2020 (Spotlight presentation: 280/9454 submissions). [Short Talk], [Long Talk], [Poster]

• Online Continuous DR-Submodular Maximization with Long-Term Budget Constraints,
  With Maryam Fazel, AISTATS 2020. [Talk], [Poster]

• Online Algorithms for Budget-Constrained DR-Submodular Maximization,
  With Reza Eghbali and Maryam Fazel, ICML 2020 workshop on Negative Dependence and Submodularity for ML (NDSML 2020). [Slides], [Full Version]

Notes and Surveys

• Data-Dependent Regret Bounds for Bandits Problems,
  With Max Gray and Tanner Fiez.

• Online Adversarial Zero-Sum Games.

• Introduction to Spectral Graph Theory,
  With Catherine Babecki and Kevin Liu.

• Linear Regression and Linear Sequential Experimental Design,
  With Johannes Linder, Felix Leeb, and Sumit Mukherjee.

• Analysis of Alternatives to Stochastic Gradient Descent.

CV

Download here.

Contact

omids@uw.edu