Samet Oymak

Assistant Professor
Electrical and Computer Eng, UC Riverside
Cooperating Faculty with Computer Science

Office: Multidisciplinary Research Building 169 Aberdeen Dr, Riverside, CA 92507
Email: lastname (at) ece (dot) ucr (dot) edu

PhD in Electrical Engineering, California Institute of Technology, 2015

SODA Lab advances statistics and optimization for data science and AI.

UPDATE: Our group will be moving to the ECE Department at the University of Michigan in Fall 2023!

Research focus: Statistical learning meets (non)convex optimization to enable resource efficient AI/ML

Current topics: deep learning theory, resource efficiency, learning & control, transformers and attention

PhD recruiting: I will be recruiting at least 2 students for Fall 2023 at the University of Michigan.
Our lab's focus is principled and empirically-impactful ML methods (see criteria below). Please drop an email!

Self-motivated and intellectually curious with how things (e.g. AI/ML systems) work
Strong background in applied math (stats, optimization) and machine learning methods
Research experience in ML theory OR innovative AI/ML algorithms (optimization, architectures, NLP, RL …)

News

02/2023 Our new paper interprets transformers as learning algorithms and characterizes their generalization and stability.
02/2023 Two papers are accepted to AAAI23. Congrats to collaborators and Yingcong!
- Y. Li and S. Oymak, “Provable Pathways: Learning Multiple Tasks over Multiple Paths”.
- Y. Qin, Y. Li, F. Pasqualetti, M. Fazel, S. Oymak, “Stochastic Contextual Bandits with Long Horizon Rewards”.
11/2022 I am giving seminars at Google NY, KTH, and UCR on our recent results.
09/2022 We welcome Emrullah to our team!
07/2022 Nael, Khaled and I received an NSF medium award on approximate computing for ML security
07/2022 Congrats to Mingchen on receiving DYP fellowship from UCR graduate division!
07/2022 Two papers are accepted to IEEE CDC 2022 and Open Journal of Control Systems
05/2022 FedNest is accepted to ICML’22 as oral presentation
05/2022 Our paper Non-Stationary Representation Learning in Sequential Linear Bandits is accepted to IEEE Open Journal of Control Systems
04/2022 Received a Google Research Scholar award!
03/2022 Karthik will be joining our lab and MURI team as a postdoctoral scholar!
03/2022 New paper on Provable and Efficient Continual Representation Learning.
02/2022 Three papers accepted to ACC 2022.
02/2022 New paper on “FedNest: Federated Bilevel, Minimax, and Compositional Optimization”.
01/2022 Our paper on “Non-asymptotic and accurate learning of nonlinear dynamical systems” is accepted to JMLR.
01/2022 ML workshop at RoseHack 2021 on “Challenges & Opportunities in Machine Learning”.
01/2022 New preprint “Non-Stationary Representation Learning in Sequential Linear Bandits”.
12/2021 Seminar on “Principles of Efficient & Fair Learning with Overparameterization” at the USC ML Symposium.
11/2021 New preprint on Adaptive Control of Markov Jump Systems.
10/2021 3 papers are accepted to NeurIPS 2021! Congrats to my students Mingchen, Xuechen, Ibrahim and all collaborators!
- ‘‘AutoBalance: Optimized Loss Functions for Imbalanced Data’’, M. Li, X. Zhang, C. Thrampoulidis, J. Chen, S. Oymak.
- ‘‘Label-Imbalanced and Group-Sensitive Classification under Overparameterization’’, GR Kini, O. Paraskevas, S. Oymak, C. Thrampoulidis.
- ‘‘Towards Sample-Efficient Overparameterized Meta-Learning’’, Y. Sun, A. Narang, HI Gulluk, S. Oymak, M. Fazel.
Fall 2021 Excited about my new course: “EE/CS 248: Optimization for Machine Learning”
09/2021 Our new work extends confidence calibration to general performance metrics and new application settings.
08/2021 Our paper on Augmenting Geographically Weighted Regression is accepted to SIGSPATIAL as a full paper.
08/2021 Seminar on our work on architecture search at Stanford Statistics.
08/2021 Our note on super-convergence with large cyclical learning rates is accepted to IEEE Signal Processing Letters!
07/2021 Our MURI proposal on Understanding neuro-glial dynamics for robust non-Markovian learning is funded!
06/2021 Three papers will be presented at RL Theory and Overparameterization Workshops at ICML.
05/2021 I am giving seminars on our recent work at EPFL, Uppsala University, and University of Iowa.
05/2021 New preprint on Certainty Equivalent Quadratic Control for Markov Jump Systems.
04/2021 Our paper on Generalization Guarantees for Neural Architecture Search will appear at ICML 2021.
04/2021 Three papers are presented in TOPML workshop.
04/2021 Our paper on Unsupervised Domain Adaptation is now available (CVPR 2021 oral).
02/2021 New preprint on learning from imbalanced data with Chris’ group.
01/2021 Our paper on RL chatbots with Hristidis group received the Best Student Paper Award at ICSC 2021.
01/2021 Received the NSF CAREER award!
01/2021 Our paper on semisupervised learning theory is accepted to AISTATS 2021.
01/2021 Two paper accepted to ICASSP 2021.
12/2020 Our paper on Provable Benefits of Overparameterization in Model Compression is accepted to AAAI 2021!
11/2020 Our paper on Matrix Profile is accepted to Machine Learning for Systems Workshop at NeurIPS 2020.
09/2020 Our paper on multiclass learning (with Chris and Mahdi) is accepted to NeurIPS 2020!
08/2020 Received UCR's Regents Faculty Fellowship.
08/2020 Our paper with Abu and Vagelis appeared at KDD.
06/2020 New preprint on pruning neural networks with Mingchen, Yahya, Chris.
05/2020 Our system identification work with Yue and Maryam is accepted to L4DC as an oral presentation!
04/2020 Our paper on neural net optimization theory is accepted to IEEE Journal on Selected Areas in Info Theory.
03/2020 Paper accepted to ICDCS in collaboration with Jiasi and Srikanth.
02/2020 New paper on calibrating heterogenous datasets.
01/2020 Our paper is accepted to AISTATS.
01/2020 Our paper is accepted to Transactions on Signal Processing.
01/2020 NSF funded our proposal on Cyber-Physical System Safety!