Distributed Zeroth-Order Optimization Professors and Supervisors

Meisam RazaviyaynView profile

Associate Professor

Swiss Federal Institute of Technology in LausanneSwitzerland

Meisam Razaviyayn is an Associate Professor at the University of Southern California (USC) with joint appointments in the Department of Industrial and Systems Engineering , Computer Science , Quantitative and Computational Biology , and Electrical Engineering . He serves as the Associate Director of the USC-Meta Center for Research and Education in AI and Learning and holds the Andrew and Erna Viterbi Early Career Chair . His academic background includes a Ph.D. in Electrical Engineering with a minor in Computer Science from the University of Minnesota, followed by a postdoctoral fellowship at Stanford University. Razaviyayn maintains active collaborations with industry partners like Google Research and Amazon. Current Roles : Faculty Visitor at Google Research Editorial Roles : Associate Editor for SIAM Journal on Optimization and IEEE Transactions on Signal Processing His research focuses on optimization algorithms for modern data science problems, particularly addressing privacy , fairness , and scalability in machine learning systems. Recent work includes differentally private optimization frameworks and fair learning algorithms applicable to large-scale and decentralized data scenarios. Key scientific awards include: 2022 NSF CAREER Award 2022 Northrop Grumman Excellence in Teaching Award 2021 AFOSR Young Investigator Award 2021 3M Nontenured Faculty Award 2020 ICCM Best Paper Award in Mathematics 2019 IEEE Data Science Workshop Best Paper Award He actively advises Ph.D. students and leads research initiatives in robust and private machine learning . His group has received funding from NSF, Google Research, and Amazon for projects related to generative AI training , differentially private algorithms , and distributed optimization . Razaviyayn also contributes to academic service as a conference chair and workshop organizer.

Official University Profile

Tommaso CesariView profile

Assistant Professor

University of OttawaCanada

Tommaso Cesari is an Assistant Professor at the School of Electrical Engineering and Computer Science and cross-appointed with the Department of Mathematics and Statistics at the University of Ottawa. He holds a Ph.D. in Computer Science (with distinction) and Master’s/Bachelor’s degrees in Mathematics (summa cum laude) from the University of Milan. His research focuses on theoretical machine learning, online optimization, and algorithmic game theory with applications to digital markets and space operations. Education: Bachelor's Degree in Mathematics (summa cum laude), Università degli Studi di Milano (2011) Master's Degree in Mathematics (summa cum laude), Università degli Studi di Milano (2015) Ph.D. in Computer Science (with distinction), Università degli Studi di Milano (2020) Tommaso’s research spans online learning , regret minimization , and optimization under limited feedback . His work bridges machine learning , game theory , and stochastic optimization , with applications to bilateral trade , market microstructure , and spacecraft operations . His recent publications emphasize regret analysis in contextual brokerage, market making , and fair bilateral trade . He has contributed to top venues like COLT , NeurIPS , EC , and MOR , often collaborating with researchers such as Nicolò Cesa-Bianchi and François Bachoc. Tommaso currently teaches ITI 1121 — Introduction to Computing II and CSI 4145 — Machine Learning at the University of Ottawa. He is a member of the ELLIS society and serves as an editor for Springer's Machine Learning Journal .

Official University Profile

Lorenzo RosascoView profile

Professor

University of GenoaItaly

Lorenzo Rosasco is a Full Professor at the Department of Computer Science, Bioengineering, Robotics and Systems Engineering (DIBRIS) at the University of Genoa. He holds roles including University contact for the thematic area "Artificial Intelligence" and membership on the Department Board. His research focuses on machine learning theory, kernel methods, optimization, robotics, and their applications in control systems, domain adaptation, and interdisciplinary fields like biomedical engineering. His educational background includes advanced expertise in mathematical foundations of machine learning, though specific academic credentials are not detailed here. Teaching responsibilities include courses such as MACHINE LEARNING, SEQUENTIAL PREDICTION AND REINFORCEMENT LEARNING, and INFORMATION THEORY AND INFERENCE at both undergraduate and graduate levels. Research interests revolve around developing efficient algorithms for large-scale machine learning, including iterative regularization, transfer learning, and control systems using Koopman operator theory. Notable contributions include work on fast kernel methods, optimization in neural networks, and applications in robotics for tasks like humanoid locomotion and tactile sensing. Recent publications explore topics such as double descent phenomena, sim-to-real transfer learning, and adaptive control strategies. His work bridges theoretical insights with practical applications, emphasizing computational efficiency and interdisciplinary impact.

Official University Profile

Emiliano Dall’AneseView profile

Associate Professor

Boston UniversityUnited States

Emiliano Dall’Anese is an Associate Professor in the Department of Electrical and Computer Engineering at Boston University, appointed in the Division of Systems Engineering. He holds affiliations with the Center for Information and Systems Engineering and the Institute for Global Sustainability. He received his Ph.D. in Information Engineering from the University of Padova, Italy, in 2011, and was a postdoctoral researcher at the University of Minnesota until 2014. Prior to BU, he was at the University of Colorado Boulder and the National Renewable Energy Laboratory (NREL). His research focuses on optimization, control, and learning in cyber-physical systems, emphasizing applications in power systems, electrified transportation, and autonomous systems. Recent funded projects include Air Force Office of Scientific Research grants for feedback controllers in stochastic environments and U.S. Department of Energy initiatives for resilient transmission systems. Key achievements include the NSF CAREER Award (2020), IEEE Transactions on Control of Network Systems Best Paper Award (2023), and multiple IEEE PES best paper awards. His work bridges foundational theory with practical applications, with over 100 publications in top journals and conferences. He advises a dynamic group of Ph.D. students and postdocs, collaborates internationally, and teaches advanced courses on control systems, optimization, and power grids. His research group actively engages in experimental validation and technology transfer for sustainable energy systems.

Official University Profile

Dan AlistarhView profile

Professor

Massachusetts Institute of TechnologyUnited States

Dan Alistarh is a Professor at the Institute of Science and Technology Austria (ISTA) and ML Research Lead at Neural Magic, Inc. His research focuses on efficient algorithms and systems for machine learning, spanning theoretical foundations to practical implementations. He holds a PhD from EPFL and has held postdoctoral positions at MIT CSAIL and researcher roles at ETH Zurich and Microsoft Research. Education: PhD in Computer Science from EPFL (advisor: Prof. Rachid Guerraoui), Postdoctoral Associate at MIT CSAIL (advisor: Prof. Nir Shavit). Research interests include distributed systems, optimization algorithms, neural network compression (quantization, pruning), and parallel computing. His work bridges theory and practice, with impactful contributions to LLM efficiency, such as GPTQ, Marlin, and SparseGPT. He leads a lab at ISTA with active collaborations and visiting researchers from top institutions. Key scientific awards include the Best Paper Award at DISC 2021 (for work on leader election algorithms) and a Distinguished Paper Award at SPAA 2023. His research is supported by grants from FWF BILAI, ERC, NVIDIA, Google, and Amazon. Advising and grants: Supervises PhD students and postdocs focusing on ML efficiency. Open positions exist for interns, PhD candidates, and postdocs with backgrounds in CS/Math. Recent lab achievements include 4 NeurIPS 2024 acceptances, including an oral presentation for PV-Tuning. Labs/Teams: IST-DASLab, known for open-source contributions (GitHub: IST-DASLab) and innovations in LLM compression. Collaborations span academia and industry, with alumni progressing to roles at OpenAI, Neural Magic, and top universities.

Official University Profile

He BaiView profile

Associate Professor

Oklahoma State UniversityUnited States

He Bai is an Associate Professor in the Department of Mechanical and Aerospace Engineering at Oklahoma State University (OSU), part of the College of Engineering, Architecture and Technology (CEAT). His research focuses on unmanned systems in GPS-denied environments, multi-agent systems, nonlinear control, atmospheric turbulence modeling, and human-robot interaction. He advises OSU's Mercury Robotics Society and Automation Society and serves as a faculty adviser for graduate and undergraduate students. Education includes a Ph.D. (2009), M.S. (2007), and B.E. (2005) in Electrical Engineering and Automation from Rensselaer Polytechnic Institute and the University of Science and Technology of China. Research interests emphasize robust control strategies for autonomous systems, including cooperative navigation, adaptive control under disturbances, and machine learning applications in robotics. His work addresses challenges like wind estimation for unmanned aerial systems (UAS) and human-vehicle collaboration in mixed traffic environments. Professional roles include Associate Editor for IEEE Control System Letters and the IEEE Control System Society Conferences. Recent publications explore reinforcement learning for multi-agent coordination, decentralized Bayesian learning, and turbulence impact on UAS navigation. Advising and lab involvement includes directing the Mercury Robotics Society and collaborating on projects like the intelligent cooperative driving testbed. His research bridges theoretical control systems with practical applications in autonomous systems and robotics.

Official University Profile

Stefan M. WildView profile

Adjunct Professor

Northwestern UniversityUnited States

Stefan M. Wild serves as Director of the Applied Mathematics and Computational Research (AMCR) Division and Senior Scientist at Lawrence Berkeley National Laboratory, while holding an adjunct faculty position in the Industrial Engineering and Management Sciences (IEMS) department at Northwestern University's McCormick School of Engineering. He also serves as a Senior Fellow at NAISE (Northwestern Initiative for AI and Society). Dr. Wild earned his Ph.D. and M.S. in Operations Research from Cornell University (2009, 2007) and his M.S. and B.S. in Applied Mathematics from the University of Colorado, Boulder (2003, 2002). His academic journey includes an Argonne Director's Postdoctoral Fellowship (2008-2010) and the DOE Computational Science Graduate Fellowship (2005-2008). His research program focuses on developing numerical optimization and automated learning algorithms for challenging science and engineering problems at interfaces involving computer simulations, complex data, and physical experiments. Dr. Wild leads multiple community software projects including BAND, parMOO, libEnsemble, deepHyper, NUCLEI, POptUS, and surmise, with applications spanning nuclear physics, materials science, and astrophysics. His work bridges derivative-free optimization, uncertainty quantification, high-performance computing, and scientific machine learning. Dr. Wild has advised numerous postdocs who have established successful careers at national laboratories and academic institutions. His editorial responsibilities include Mathematical Programming Computation, INFORMS Journal on Computing, Data Science in Science, and SIAM Review. U.S. Department of Energy Early Career Research Award (2020) STS Forum Future Leader (2018) IDC HPC Innovation Excellence Award (2015) SIAM SIGEST Award (2014) Strategic Laboratory Leadership Program (UChicago Booth School) (2013) DOE Computational Science Graduate Fellowship (2005-2008) As AMCR Division Director, he leads a diverse team of applied mathematicians, computational scientists, and software engineers addressing some of the world's most challenging computational problems across scientific and engineering disciplines. His leadership emphasizes both technical excellence and commitment to inclusion, diversity, equity, and accountability in scientific research.

Official University Profile

Sébastien Le DigabelView profile

Professor

Polytechnic School of MontrealCanada

Sébastien Le Digabel is a Full Professor in the Department of Mathematics and Industrial Engineering at Polytechnique Montréal. He holds affiliations with GERAD (Groupe d'études et de recherche en analyse des décisions) and IVADO (Institut de valorisation des données), where he contributes to research in optimization and artificial intelligence. Professor Le Digabel's research focuses on black-box optimization and derivative-free methods, particularly the Mesh Adaptive Direct Search (MADS) algorithm. His work spans three main areas: development of optimization methods, application to engineering problems, and development of optimization software. He has made significant contributions to extending MADS for constrained optimization, mixed-variable problems, multiobjective optimization, and stochastic optimization. His recent publications (2023-2025) demonstrate continued innovation in handling complex optimization challenges including categorical variables, hierarchical constraints, and multi-fidelity approaches. These works appear in top journals like Computational Optimization and Applications, SIAM Journal on Optimization, and Optimization and Engineering. Professor Le Digabel actively supervises graduate students, having completed 8 PhD theses and 10 Master's theses with topics spanning multiobjective optimization, hyperparameter tuning, smart grid applications, and power systems optimization. His research is supported by diverse funding sources including NSERC, FRQNT, InnovÉÉ, IVADO, Hydro-Québec, Rio Tinto, and Huawei-Canada, reflecting the broad applicability of his work across academic and industrial domains. Professor Le Digabel leads the NOMAD Research Group and is the primary developer of the NOMAD optimization software, which implements the MADS algorithm and serves as a standard tool in the derivative-free optimization community. The software is available at www.gerad.ca/nomad and continues to evolve with new capabilities for challenging optimization problems.

Official University Profile