جست‌وجوی استاد و پژوهشگر

• Deep Learning
• Distributed Machine Learning
• Federated Learning
• +۷ مورد دیگر

Tao Lin is a Tenure-Track Assistant Professor and Principal Investigator of LINs Lab at Westlake University, School of Engineering. He leads cutting-edge research in deep learning optimization, generalization, and robustness, particularly in distributed and federated settings. Prior to this, he was a Ph.D. student at École Polytechnique Fédérale de Lausanne (EPFL), Switzerland, under the supervision of Prof. Martin Jaggi and Prof. Babak Falsafi. Doctor of Science, School of Computer and Communication Sciences, EPFL, Switzerland (2017–2022) Master of Science, School of Computer and Communication Sciences, EPFL, Switzerland (2014–2017) Bachelor of Engineering (with honors), College of Electrical Engineering, Zhejiang University, China (2010–2014) His research focuses on the intersection of optimization and generalization in deep learning, leveraging theoretical and empirical insights into loss landscapes and training dynamics to design efficient and robust learning and inference methods. This includes work on decentralized and federated learning under noisy, heterogeneous, and hardware-constrained environments. His work spans algorithmic innovation, theoretical analysis, and practical system integration. The recent publications from his lab demonstrate a strong trend in advancing federated learning, efficient inference for large language models, multimodal foundation models in pathology, and robust training under distribution shifts. Key themes include communication efficiency, model personalization, gradient tracking, and hardware-aware learning. His group has published at top venues including NeurIPS, ICML, ICLR, CVPR, and ECCV, with several papers receiving oral or spotlight presentations. ECCV Best Paper Candidate, 2024 Top 2% Scientists Worldwide 2024 (Stanford University) Doctoral Program Thesis Distinction Award, EPFL, 2022 Outstanding Performance Bonus, EPFL, 2021–2022 Top Reviewer: NeurIPS, ICML, AISTATS He advises multiple Ph.D. and master’s students, including Yongxin Guo, Futing Wang, Peng Sun, and Yuxuan Sun, whose work has been accepted at premier conferences. He has secured competitive grants as PI and participant, including the National Natural Science Foundation of China for Excellent Young Scientists Fund (Overseas) and the Science and Technology Innovation 2030 – Major Project. He also contributes to the community through service as an area chair (NeurIPS, ICML), reviewer for top journals and conferences, and organizer of workshops and academic events. His open-source contributions, such as Post-local SGD, have been integrated into PyTorch. Tao Lin teaches graduate courses such as Research Methodology of Computer Science and Technology and Deep Learning at Westlake University. He is actively involved in academic governance, serving on committees for student seminars, academic exchange, doctoral studies, and teaching leadership. The LINs Lab runs a regular research seminar on Deep Learning and Optimization, fostering a collaborative and dynamic research environment.

Jean-Yves Le Boudecمشاهده پروفایل

استاد

• Network Calculus
• Time-Sensitive Networking
• Deterministic Networking
• +۷ مورد دیگر

Jean-Yves Le Boudec is a Professor at École Polytechnique Fédérale de Lausanne (EPFL), affiliated with the School of Computer and Communication Sciences and the Institute of Electrical Engineering. He has been a key figure in advancing the theory and application of network calculus and deterministic networking, contributing significantly to standards such as IEEE Time-Sensitive Networking (TSN) and IETF DetNet. His research focuses on network calculus , time-sensitive and deterministic networking , traffic regulation , worst-case delay analysis , and cyber-physical systems , with cross-cutting applications in smart grids , real-time communication , and network security . He has co-authored foundational texts on network calculus and developed theoretical frameworks for traffic regulators, service curves, and delay bounds in complex networked systems. The recent publications highlight a strong trend in analyzing and improving performance guarantees in deterministic networks, including scheduling mechanisms like Deficit Round-Robin and Cyclic Queuing and Forwarding, traffic shaping via interleaved regulators, and security against time-synchronization attacks in power systems. The work spans theoretical modeling using stochastic and min-plus/max-plus algebra, practical algorithm design, and application to critical infrastructure. IEEE Fellow Le Boudec has advised numerous researchers and PhD students, including Ehsan Mohammadpour, Ludovic Thomas, and Seyed Mohammadhossein Tabatabaee. His collaborative projects often involve grants related to European and Swiss research initiatives in networking and smart grid technologies. He leads a research group focused on networked systems at EPFL, contributing to both theoretical advances and real-world implementations in industrial and energy-critical networks. His lab work centers on modeling and verification of time-sensitive network behaviors, integrating formal methods with practical experimentation. The team investigates regulators, shapers, and synchronization mechanisms, aiming to ensure robustness, predictability, and security in next-generation communication infrastructures. Future work continues to explore the interplay between communication, control, and energy systems in highly reliable environments.

• Applied Analysis
• Numerical Analysis
• Data Science
• +۶ مورد دیگر

Leon Bungert is a Professor of Mathematics of Machine Learning at the University of Würzburg, working in applied analysis and numerics with a particular focus on data science and machine learning. His research investigates PDEs and variational models on graphs, adversarial robustness of machine learning, variational regularization, and nonlinear optimization. Dr. Bungert serves as a guest editor for the European Journal of Applied Mathematics, an associate editor for Advances in Continuous and Discrete Models: Theory and Applications, and is a member of the program committee at SSVM 2025. He is also an ELLIS member and actively organizes conferences and workshops, including "MIA'25" at IHP in Paris (January 13-15, 2025), "Synergies of Machine Learning and Numerics" in Osaka (March 11-13, 2025), and "Mathematical Analysis of Adversarial Machine Learning" in Oaxaca (August 17-22, 2025). Research Interests Dr. Bungert's primary research areas include: PDEs on graphs Adversarial robustness in machine learning Inverse problems Optimization Variational problems in L-infinity Nonlinear eigenvalue problems Image reconstruction with structural priors His work bridges theoretical mathematics with practical applications in machine learning, particularly focusing on the mathematical foundations of deep learning and developing robust algorithms that can withstand adversarial attacks. He has made significant contributions to understanding the connections between partial differential equations and machine learning algorithms. Research Trends Analysis of Dr. Bungert's recent publications reveals a strong focus on the intersection of machine learning and mathematical analysis. A key theme is the application of variational methods and partial differential equations to machine learning problems, particularly in understanding and improving the robustness of neural networks against adversarial examples. His work on Lipschitz learning on graphs has established important theoretical foundations for graph-based semi-supervised learning. Additionally, his research on the infinity Laplacian and p-Laplacian equations provides deep insights into the mathematical structure of machine learning algorithms. The development of Bregman learning frameworks for sparse neural networks represents a significant contribution to efficient deep learning model training. Professional Activities Dr. Bungert is actively involved in the academic community through editorial roles and conference organization. His current professional activities include: Guest editor for the European Journal of Applied Mathematics Associate editor for Advances in Continuous and Discrete Models: Theory and Applications Member of the program committee at SSVM 2025 ELLIS member Co-organizer of multiple international conferences and workshops Technical Contributions Dr. Bungert has developed several open-source software packages that implement his theoretical contributions, including: Code for convergence rates of Lipschitz learning on graphs A Bregman training framework for sparse neural networks CLIP: Cheap Lipschitz Training of Neural Networks Nonlinear Power Method for Proximal Operators and Neural Networks Robust Image Reconstruction with Misaligned Structural Information These implementations are primarily in Python and MATLAB, demonstrating his commitment to making theoretical advances accessible for practical applications.

Franck Iutzelerمشاهده پروفایل

استاد

• Numerical optimization
• Statistical learning
• Stochastic programming
• +۵ مورد دیگر

Franck Iutzeler is a Professor of Applied Mathematics at Université de Toulouse, working within the Statistics & Optimization team of the Institut Mathématique de Toulouse and teaching in the Department of Mathematics. He previously served as an Assistant Professor at Université Grenoble Alpes from 2015 to 2023 and completed his Habilitation à Diriger des Recherches in 2021. His research focuses on the intersection of optimization, statistics, and optimal transport theory to develop robust data-driven models. Key areas include numerical optimization, statistical learning, stochastic programming, and optimal transport. He is particularly interested in distributionally robust optimization using Wasserstein metrics and has developed the skwdro Python library for implementing these methods. Iutzeler's recent publications demonstrate a strong focus on Wasserstein Distributionally Robust Optimization (WDRO), with multiple papers in top venues like NeurIPS and SIAM Journal on Optimization. His work bridges theoretical guarantees with practical implementation, particularly through the skwdro library which provides efficient code for WDRO in machine learning applications. ANR JCJC grant for project STROLL: Harnessing Structure in Optimization for Large-scale Learning Co-PI of ANITI chair on Trust and Responsibility in Artificial Intelligence led by JM. Loubes and J. Bolte Iutzeler actively supervises PhD students including Yu-Guan Hsieh (awarded Université Grenoble Alpes's PhD award), Gilles Bareilles, Waïss Azizian, and Victor Mercklé. He has secured research funding through the ANR (MAD project on Automatic Differentiation) and ANITI. His current research includes statistical fairness using optimal transport theory and automatic differentiation for stochastic optimization. He leads the development of the skwdro library for Wasserstein Distributionally Robust Optimization and is involved with ANITI (Toulouse's AI Cluster), where he also took responsibility for the 2nd year of the Master SID in Data Science & Engineering in September 2024.

Helmut Bölcskeiمشاهده پروفایل

استاد

• Applied Mathematics
• Machine Learning Theory
• Mathematical Signal Processing
• +۳ مورد دیگر

Prof. Dr. Helmut Bölcskei is a Full Professor of Mathematical Information Science at ETH Zurich's Department of Information Technology and Electrical Engineering. He holds a joint affiliation with the Department of Mathematics. His academic journey includes a Dipl.-Ing. and Dr. techn. from Vienna University of Technology, followed by postdoctoral research at Stanford University and industry roles at Iospan Wireless and Celestrius AG. He has been at ETH Zurich since 2002, contributing to applied mathematics, machine learning theory, signal processing, and statistics. Education : 1994: Dipl.-Ing., Vienna University of Technology 1997: Dr. techn., Vienna University of Technology Industry Experience : Co-founder of Iospan Wireless (acquired by Intel) and Celestrius AG His research focuses on applied mathematics , machine learning theory , and data science , with emphasis on neural network approximation, metric entropy, and signal processing. Recent work explores theoretical limits of deep learning and nonlinear system identification. His publications highlight advancements in quantization, compression, and system complexity analysis. Prof. Bölcskei has received numerous accolades, including IEEE Fellow status, the 2010 Vodafone Innovations Award, and the ETH 'Golden Owl' Teaching Award. He served as Editor-in-Chief of the IEEE Transactions on Information Theory (2010–2013) and has held editorial roles in multiple journals. His leadership includes roles on the Board of Governors of the IEEE Information Theory Society and as a delegate for faculty appointments at ETH Zurich. Labs/Teams : Mathematical Information Science Group at ETH Zurich's Department of Information Technology and Electrical Engineering

Alin Achimمشاهده پروفایل

استاد

• Statistical Signal Processing
• Inverse Problems
• Computational Imaging
• +۷ مورد دیگر

Alin Achim is a Professor of Computational Imaging at the School of Computer Science, University of Bristol, with a focus on statistical signal processing and inverse problems. His research bridges model-based and data-driven approaches in machine learning, particularly for computational imaging applications. Education: B.Sc. and M.Sc. from the University of Bucharest, Ph.D. from the University of Patras. His work emphasizes sparse representations in signal/image processing, enabling non-linear algorithms for image denoising, fusion, segmentation, super-resolution , and novel penalty functions for deep learning architectures. Applications span Synthetic Aperture Radar (SAR) imaging for Earth Observation and biomedical imaging (e.g., lung ultrasound, OCT). Recent publications highlight trends in computational medical imaging (2023), SAR for ocean surface modeling (2022), and non-convex optimization methods (2020). His research integrates statistical sparsity with classical sparsity constraints. Scientific Awards: EUSIPCO Best Student Paper Award (2009) Senior Member IEEE (2009) He has supervised multiple students and leads active projects like Next Generation Quantitative Acoustic Microscopy (NIH-funded) and AssenSAR for naval SAR applications. He contributes to open-access datasets (e.g., OCT2Confocal, 3D OCT Images of Murine Uveitis) and serves on editorial boards for journals like IEEE Transactions on Image Processing .

• Online learning
• Optimization theory
• Bandits
• +۲ مورد دیگر

Tommaso Cesari serves as an Assistant Professor in the EECS department at the University of Ottawa, cross-appointed with the Department of Mathematics and Statistics. He currently teaches Machine Learning (CSI 4145) and Introduction to Computing II (ITI 1121 A), courses he has instructed since 2024 and 2023 respectively. His research centers on theoretical machine learning with applications in market design and operational systems. Key interests include online learning frameworks for bandit problems, pricing mechanisms, bilateral trade optimization, and spacecraft operations scheduling. This work bridges algorithmic theory with real-world economic systems, emphasizing regret minimization under adversarial conditions and smoothed analysis. Analysis of his 2023-2025 publications reveals a cohesive trajectory in applying online learning to bilateral trade mechanisms (appearing in 7 of 15 recent papers), auction theory, and brokerage models. His work consistently targets top venues including JMLR, NeurIPS, and STOC, with growing emphasis on fairness constraints and multi-platform systems in 2024-2025 outputs. Professor Cesari actively recruits Master's and Ph.D. students for research in his domains; prospective candidates may apply through the university portal. His collaborative publications with prominent researchers like Nicolò Cesa-Bianchi indicate strong engagement in the theoretical machine learning community.

• Positional games
• Discrete and computational geometry
• Discrete random structures
• +۲ مورد دیگر

Miloš Stojaković is a Full Professor at the Department of Mathematics and Informatics, Faculty of Sciences, University of Novi Sad, Serbia. He has held this position since 2016, following roles as Associate Professor (2011–2016) and Assistant Professor (2006–2011). His research focuses on positional games, discrete and computational geometry, discrete random structures, combinatorial algorithms, and graph theory. He leads the Foundations of Computer Science group since 2018. Stojaković earned his Ph.D. in Computer Science from ETH Zurich (2005), advised by Emo Welzl and Tibor Szabó, and holds M.Sc. and B.Sc. degrees from the University of Novi Sad. He has been recognized with the Dr Z. Đinđić Award (2008) for best young scientist in Vojvodina and the Best Student of University of Novi Sad Award (1998/99). He has advised three Ph.D. students: Mirjana Mikalački, Marko Savić, and Jelena Stratijev. His work spans over 60 publications, including seminal contributions to positional games and computational geometry. He serves on editorial boards of Discrete Mathematics & Theoretical Computer Science and the Novi Sad Journal of Mathematics . Stojaković teaches courses such as Combinatorial Algorithms, Graph Theory, and Theoretical Computer Science at the University of Novi Sad. He has also taught specialized courses on positional games at institutions like the University of Buenos Aires and Eötvös Loránd University Budapest. His research interests emphasize algorithmic and combinatorial aspects of games, geometry, and graph theory.

• Computer Science
• Machine Learning
• Artificial Intelligence
• +۳ مورد دیگر

Yongtao Wu is a Doctoral Assistant and student in the doctoral program in Computer and Communication Sciences at École Polytechnique Fédérale de Lausanne (EPFL). He is affiliated with the Laboratory for Information and Inference Systems (LIONS), part of the School of Engineering (STI) and the Institute of Electrical Engineering (IEM). His research focuses on machine learning, adversarial learning, computer vision, and deep learning, with applications in vision-language models, robustness analysis, and optimization algorithms. His work explores topics such as diffusion-based purification for vision-language models, adversarial attacks on large models, and parameter-efficient fine-tuning techniques. He contributes to advancing foundational AI models and their robustness in real-world scenarios. His research also intersects with cybersecurity, structural engineering predictions, and quantum computing methodologies. Yongtao is actively involved in the LIONS lab, collaborating on projects that bridge theoretical advancements with practical applications in computer science and engineering. His doctoral studies emphasize interdisciplinary approaches to solving complex problems in AI and machine learning.

• Numerical Analysis
• High-Dimensional Problems
• Low-Rank Matrix and Tensor Techniques
• +۴ مورد دیگر

Bart Vandereycken is an Associate Professor in the Mathematics Department at the University of Geneva, specializing in numerical analysis and scientific computing. His research focuses on large-scale and high-dimensional problems solved using low-rank matrix and tensor techniques, with applications in numerical linear algebra, optimization, and nonlinear eigenvalue problems. He previously held positions as an instructor at Princeton University and postdoctoral researcher at EPF Lausanne and ETH Zurich, and earned his PhD from KU Leuven in 2010. His research interests include Riemannian optimization algorithms, multilevel preconditioning, and machine learning applications. He serves as an associate editor for SIAM Journal on Matrix Analysis and Applications and Linear Algebra and its Applications . Bart organizes the Numerical Analysis seminar with colleagues, and advises students interested in numerical analysis or numerical linear algebra. Recent work emphasizes convexity structures in matrix decompositions, robust preconditioning techniques, and scalable low-rank algorithms for high-dimensional PDEs. His 2024–2025 publications explore advancements in Riemannian optimization schemes, subspace iteration methods, and distributed computing applications of matrix decompositions. Key themes include improving convergence guarantees and developing geodesic-based optimization frameworks for challenging numerical problems.

Masashi Sugiyamaمشاهده پروفایل

استاد

• statistical machine learning
• weakly supervised learning
• learning from noisy supervision
• +۶ مورد دیگر

Masashi Sugiyama is a Professor at the Department of Complexity Science and Engineering, Graduate School of Frontier Sciences, The University of Tokyo, where he has been serving since 2014. He concurrently serves as the Director of the RIKEN Center for Advanced Intelligence Project (AIP) since 2016, leading research groups focused on fundamental AI technologies, AI applications, and social issues of AI. His academic journey began at Tokyo Institute of Technology, where he earned his Bachelor, Master, and Doctor of Engineering degrees in Computer Science in 1997, 1999, and 2001 respectively, before becoming an Assistant Professor and later Associate Professor at the same institution. Sugiyama's research primarily focuses on statistical machine learning, with particular expertise in weakly supervised learning, learning from noisy supervision, and learning under distribution shift. His work has established important theoretical frameworks for density ratio estimation, covariate shift adaptation, and non-stationary environments. He has developed numerous practical algorithms including KLIEP (Kullback-Leibler importance estimation procedure), uLSIF (unconstrained least-squares importance fitting), and LSDD (least-squares density difference), which have become standard tools in the machine learning community. Analysis of his recent publications reveals a strong emphasis on reliable and robust machine learning techniques that can handle imperfect supervision and data distribution changes. His work spans theoretical foundations, algorithm development, and practical applications across various domains including bioinformatics, anomaly detection, and dimensionality reduction. The recurring themes in his research include direct density ratio estimation methods, importance weighting techniques, and mutual information-based approaches for various machine learning tasks. Award for Science and Technology from the Japanese Minister of Education, Culture, Sports, Science and Technology (2022) Japan Academy Medal (2017) Japan Society for the Promotion of Science Award (2017) Faculty Award from IBM (2007) Nagao Special Researcher Award from the Information Processing Society of Japan (2011) Young Scientists' Prize for the Commendation for Science and Technology by the Minister of Education, Culture, Sports, Science and Technology Japan (2014) Sugiyama has served as program co-chair for major machine learning conferences including NeurIPS 2015, AISTATS 2019, and ACML 2010 and 2020. He has received prestigious fellowships including the Alexander von Humboldt Foundation Research Fellowship (2003-2004) and the European Commission Program Erasmus Mundus Scholarship (2006). He is the (co-)author of influential machine learning monographs including Machine Learning in Non-Stationary Environments (MIT Press, 2012), Density Ratio Estimation in Machine Learning (Cambridge University Press, 2012), Statistical Reinforcement Learning (Chapman & Hall, 2015), and Machine Learning from Weak Supervision (MIT Press, 2022). At RIKEN AIP, Sugiyama leads research initiatives focused on developing robust AI systems that can operate reliably in real-world conditions with imperfect data. His laboratory maintains an active software development effort, providing open-source implementations of many of his research contributions, which has significantly influenced both academic research and practical applications of machine learning.

Julien Fageotمشاهده پروفایل

پژوهشگر

• Sparsity
• Random Processes
• Approximation Theory
• +۷ مورد دیگر

Julien Fageot is a Researcher at the École Polytechnique Fédérale de Lausanne (EPFL) in the AudioVisual Communications Laboratory within the School of Computer and Communication Sciences. He previously held postdoctoral positions at Harvard University, McGill University, and EPFL. His educational background includes: Ph.D. in Electrical Engineering at EPFL (2012-2017) M.Sc. Mathematics, Vision, and Learning in ENS Paris-Saclay, France (2011) M.Sc. in Probability and Statistics at Université Paris Orsay, France (2009) École Normale Supérieure, Section Mathématiques, Paris, France (2007-2012) Dr. Fageot's research lies at the intersection of high-level mathematics and data sciences, focusing on mathematical properties of advanced processing tools for sparse signal reconstruction and synthesis. His expertise spans sparsity, random processes, approximation theory, splines, convex optimization, functional analysis, and signal/image processing. He explores probability theory (sparse stochastic processes), optimization theory (sparsity-promoting spline reconstruction), and applications in signal processing (inverse problems, segmentation, detection, CNNs). His publication record demonstrates a clear progression from theoretical foundations of stochastic processes to practical applications in biomedical imaging. Recent work shows increasing focus on machine learning applications while maintaining strong mathematical rigor, particularly in developing sparse representations for medical image analysis. His scientific achievements have been recognized with: Best Paper Award at the MIDL Conference (2019) EPFL Best Doctorate Award (2018) Outstanding PhD Thesis Distinction in Electrical Engineering, EPFL (2017) Education Award from the Life Science Department, EPFL (2013) Dr. Fageot actively mentors students, currently supervising PhD candidate Adrian Jarret and having previously guided Thomas Debarre and Shayan Aziznejad to completion. He has supervised numerous master's theses on spline-based reconstruction and biomedical image analysis. His research is supported by Swiss National Science Foundation grants including the Postdoc.Mobility fellowship for 'Mathematical Models for Analog Data Sciences: the Continuous Way' (2020) and the Early Postdoc.Mobility fellowship for 'Probabilistic and Variational Methods for Sparse Signals' (2018). As a key member of EPFL's AudioVisual Communications Laboratory, he collaborates with Prof. Martin Vetterli, Prof. Michael Unser, and Prof. Christian Genest, bridging theoretical mathematics with practical applications in signal processing and data science.

Quentin Denoyelleمشاهده پروفایل

دانشیار

• Sparse Inverse Problems
• Convex Optimization Algorithms
• Machine Learning
• +۲ مورد دیگر

Quentin Denoyelle is an Associate Professor at MAP5 (CNRS UMR 8145) at Université Paris Cité. Prior to this, he conducted postdoctoral research at the Biomedical Imaging Group (BIG) of Michael Unser at EPFL and earned his PhD in Applied Mathematics under Gabriel Peyré and Vincent Duval at CEREMADE, Université Paris-Dauphine. His research focuses on sparse inverse problems convex optimization algorithms machine learning applications signal and image processing . He is based at Université Paris Cité, working within the MAP5 laboratory, a CNRS joint research unit (UMR 8145).

• Positional games
• Discrete geometry
• Computational geometry
• +۳ مورد دیگر

Miloš Stojaković is a Full Professor at the Department of Mathematics and Informatics, Faculty of Sciences, University of Novi Sad, Serbia. He has been with the university since 1999, progressing from student to Assistant Professor (2006-2011), Associate Professor (2011-2016), and Full Professor (2016-present). He leads the Foundations of Computer Science group and serves as an editor for Discrete Mathematics & Theoretical Computer Science and the Novi Sad Journal of Mathematics. Ph.D. in Computer Science, ETH Zurich (2005), advisors: Emo Welzl and Tibor Szabó M.Sc. in Computer Science, University of Novi Sad (2001), advisor: Joviša Žunić B.Sc. in Mathematics and Computer Science, University of Novi Sad (1999) Stojaković's research centers on positional games, discrete and computational geometry, discrete random structures, combinatorial algorithms, and graph theory. His work explores theoretical aspects of combinatorial games, particularly Maker-Breaker type games on graphs and other structures. He investigates geometric properties of discrete structures and develops algorithms for geometric problems. His research has applications in theoretical computer science and combinatorics, with significant contributions to understanding game strategies, structural properties, and computational complexity in discrete settings. Stojaković's recent publications (2016-2025) primarily focus on positional games, extending classical results to new settings like posets, generalized saturation, and constructor-blocker variants. He examines computational complexity aspects and structural properties of game outcomes. His work often involves collaborations with leading researchers including Dan Hefetz, Michael Krivelevich, and Tibor Szabó, demonstrating the ongoing vitality of combinatorial game theory and its connections to other areas of discrete mathematics. "Dr Z. Đinđić Award" for the best young scientist in Vojvodina (2008) "Best Student of University of Novi Sad Award" (1998/99) Stojaković has supervised three PhD students to completion: Mirjana Mikalački (2014), Marko Savić (2018), and Jelena Stratijev (2023). His research has been supported by various grants, including the 1 million RSD grant accompanying the Dr Z. Đinđić Award. He has organized multiple workshops including the Novi Sad Workshop on Foundations of Computer Science series (2015, 2017, 2024) and co-organized the Mini-Workshop on Positional Games at Oberwolfach Research Institute (2018) and the Positional Games workshop within A Sparse (Graphs) Coalition (2021). Since 2018, Stojaković has led the Foundations of Computer Science group at the University of Novi Sad. He maintains active international collaborations, having taught courses at Eötvös Loránd University Budapest, University of Buenos Aires, University of Szeged, University of Banja Luka, University of Groningen, and Graz University of Technology. His research group focuses on theoretical aspects of combinatorial games, discrete geometry, and related computational problems, contributing significantly to these fields through publications, workshops, and academic leadership.

Zebang Shenمشاهده پروفایل

مدرس

• Optimization
• Machine Learning
• Data Science
• +۳ مورد دیگر

Dr. Zebang Shen serves as a Lecturer in the Department of Computer Science at ETH Zurich, affiliated with the Institute for Machine Learning (Institut für Maschinelles Lernen). His research activities are centered at Andreasstrasse 5, 8092 Zürich, Switzerland, with teaching responsibilities confirmed for the Autumn Semester 2025. His primary research domains include Optimization, Machine Learning, and Data Science, with specialized focus on Federated Learning, Stochastic Optimization, and Reinforcement Learning. Shen develops algorithmic solutions for projection-free optimization, minimax problems, and diffusion model applications, emphasizing theoretical guarantees alongside practical implementations in distributed learning environments. Analysis of his 2021-2025 publications reveals consistent innovation in optimization frameworks for machine learning, particularly in federated settings where privacy-utility tradeoffs and straggler resilience are addressed. His work bridges mathematical rigor (e.g., Poincaré inequalities, McKean-Vlasov equations) with scalable algorithms for real-world data science challenges. No scientific awards were documented in the available sources. While the sources confirm his faculty role and publication record, specific details regarding student advising, grant funding, or laboratory leadership were not provided. His current teaching activities indicate ongoing academic engagement at ETH Zurich. Shen operates within ETH Zurich's Institute for Machine Learning, contributing to the Department of Computer Science's research ecosystem focused on advancing machine learning theory and applications.