Find Academic Supervisors

Chen XuView profile

Associate Professor

• Sparse Modeling
• Statistical Learning
• Feature Selection
• +5 more

Chen Xu is an Associate Professor in the Department of Mathematics and Statistics at the University of Ottawa. He holds an M.A. from York University and a PhD from the University of British Columbia. His research focuses on sparse modeling, statistical learning, and big data processing, with an emphasis on both theoretical and computational advancements. Dr. Xu is affiliated with the Faculty of Science and contributes to editorial roles for journals such as the Journal of the American Statistical Association and Electronic Journal of Statistics. Education: M.A., York University PhD, University of British Columbia Research interests include feature selection, regularization methods, kernel methods, and high-dimensional regression. His work addresses computational challenges in big data, proposing efficient algorithms for tasks like singular value decomposition, clustering, and distributed feature screening. Recent publications highlight advancements in multiview PCA, low-tubal-rank tensor recovery, and model-free regression techniques. Publications span prestigious journals like the Journal of the American Statistical Association and IEEE Transactions series, focusing on statistical methodology, machine learning applications, and scalable computational solutions for complex data problems. Editorial Service: Associate Editor, Journal of American Statistical Association-T&M (2023–present) Associate Editor, Electronic Journal of Statistics (2023–present) Former Associate Editor, The Canadian Journal of Statistics (2019–2021) His research groups are Statistics and Biostatistics, and Data Science, Machine Learning, and Artificial Intelligence. He has no listed awards but maintains active editorial and academic collaborations in computational statistics and machine learning.

Soledad VillarView profile

Assistant Professor

• Optimization for Data Science
• Machine Learning
• Representation Learning
• +5 more

Soledad Villar is an Assistant Professor in the Department of Applied Mathematics and Statistics and a member of the Mathematical Institute for Data Science at Johns Hopkins University. She also contributes to the Data Science and AI Institute . Her research focuses on computational methods for extracting information from data, emphasizing optimization for data science, machine learning, equivariant representation learning, and graph neural networks. Dr. Villar holds a PhD in Mathematics from the University of Texas at Austin and has been a research fellow at New York University and the Simons Institute at UC Berkeley. Her work bridges theoretical foundations with practical applications in fields like scientific computing and political analysis. Awards include the National Science Foundation CAREER Award (2024). Her research has addressed topics such as gerrymandering detection, fluid dynamics modeling, and graph representation learning. She collaborates on interdisciplinary projects and organizes academic events like the One World MINDS Seminar and the Cibercoloquio Latinoamericano de Matemáticas . Her research interests span computational methods, equivariant machine learning frameworks, and graph neural networks, with applications in physics, engineering, and data-driven decision-making. She actively engages in advancing machine learning techniques for scientific and engineering challenges.

• GPU Computing
• Parallel Computing
• Domain-Specific Languages
• +5 more

Prof. Dr.-Ing. Richard Membarth is a Research Professor for System-on-a-Chip and AI at the Edge Computing at Technische Hochschule Ingolstadt (THI). He is affiliated with the Hardware-Software Co-Design group and holds a secondary position at the German Research Center for Artificial Intelligence (DFKI) Saarbrücken. Co-creator of DSL frameworks like AnyDSL and Hipacc Key contributor to MetaDL (AI metaprogramming) and PRIME (predictive rendering) His research bridges GPU computing , domain-specific languages , and compiler technology , with recent work on Vulkan SPIR-V compilation and device-driven SpMV algorithms . Notable awards include the HiPEAC Paper Award (2018) and multiple Best Paper Awards for his compiler frameworks.

Vivienne SzeView profile

Professor

• Energy-Efficient Machine Learning
• Computer Vision
• Video Compression/Processing
• +7 more

Vivienne Sze is a Professor at MIT's Department of Electrical Engineering and Computer Science (EECS), part of the School of Engineering. Her research focuses on energy-efficient computing systems for machine learning, computer vision, and video compression, with applications in autonomous systems, healthcare, and IoT. She leads projects integrating algorithmic innovations with hardware design to achieve low-power solutions for embedded and mobile devices. Her work has been recognized through prestigious awards, including the Primetime Engineering Emmy Award for co-developing the HEVC video compression standard and multiple faculty awards from tech giants like Google and Qualcomm. She co-authored the book *Efficient Processing of Deep Neural Networks*, emphasizing practical hardware-software co-design strategies. Research Interests: Energy-Efficient Machine Learning Accelerators Video Coding and Compression Standards Embedded Systems and Mobile Computing Processing-in-Memory (PIM) Architectures AI for Health Monitoring and Digital Health Sustainability in AI Infrastructure Publications highlight trends in: Optimizing DNNs for edge devices Innovations in entropy coding and CABAC Memory-efficient Gaussian-based algorithms Energy-aware design for photonic computing Awards include IEEE conference best paper awards and industry recognitions for her contributions to video coding and hardware acceleration. Her lab's collaborative efforts span academia and industry, aiming to bridge theoretical research with real-world deployable systems.

Grey BallardView profile

Associate Professor

• Numerical Linear Algebra
• High Performance Computing
• Computational Science
• +6 more

Dr. Grey Ballard is an Associate Professor in the Department of Computer Science at Wake Forest University . He earned a B.S. in Math and Computer Science (2006), M.A. in Math (2008) from Wake Forest, and PhD in Computer Science (2013) from the University of California, Berkeley. He was a Truman Fellow at Sandia National Laboratories before joining Wake Forest. Research Focus: Ballard develops communication-optimal algorithms for high-performance computing , particularly in tensor decompositions , symmetric matrix computations , and nonnegative matrix factorization . His work combines numerical linear algebra with parallel algorithm design to reduce data movement costs in distributed systems. Publications demonstrate expertise in communication lower bounds , randomized tensor rounding , and visualization tools for parallel algorithms. He has contributed software packages such as TuckerMPI , GentenMPI , and PLANC for large-scale data compression and clustering. Scientific Awards: Wake Forest Excellence in Research Award NSF CAREER Award SIAM Linear Algebra Prize Three Conference Best Paper Awards (SPAA, IPDPS, ICDM) C.V. Ramamoorthy Distinguished Research Award (UC Berkeley) ACM Doctoral Dissertation Award – Honorable Mention Teaching: Courses include Introduction to Computer Science , Numerical Linear Algebra , and Parallel Algorithms . He has developed educational tools using the Thread-Safe Graphics Library to visualize parallel dynamic programming and collective communication.

• Riemannian Optimization
• Matrix Manifolds
• Low-Rank Matrix Completion
• +4 more

Pierre-Antoine Absil is a Full Professor at the Louvain Polytechnic School (EPL) , part of the Université catholique de Louvain (UCLouvain) , affiliated with the Mathematical Engineering Center (INMA) and the Institute of Information and Communication Technologies, Electronics and Applied Mathematics (ICTEAM) . His research spans Riemannian optimization , matrix manifolds , and low-rank approximation with applications in astronomical imaging , signal processing , and environmental data imputation . His work emphasizes Riemannian optimization algorithms matrix and tensor completion geometric data analysis applications in astrophysics and bioinformatics Recent publications focus on Stiefel and Grassmann manifolds , variable projection methods , and exoplanet detection via alternating minimization . While no explicit scientific awards are listed, his collaborations with institutions like IEEE and Springer highlight his impact in computational mathematics and engineering. He has supervised researchers such as Simon Vary , Guillaume Olikier , and Valentin Christiaens through projects in direct imaging , economic dispatch , and graph-based data processing . His laboratory, INMA , drives methodological advances in manifold-valued data analysis and nonlinear optimization .

Saman AmarasingheView profile

Professor

• Programming Languages
• Compiler Design
• High-Performance Computing
• +5 more

Professor Saman Amarasinghe is a faculty member in the Department of Electrical Engineering and Computer Science at the Massachusetts Institute of Technology (MIT), where he leads the Commit compiler research group at MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL). His research focuses on programming languages and compilers that maximize application performance on modern computing platforms, with a particular emphasis on high-performance domain-specific languages. Professor Amarasinghe received his bachelor's degree in electrical engineering and computer science from Cornell University in 1988, followed by master's and PhD degrees in electrical engineering from Stanford University in 1990 and 1997, respectively. He joined the MIT faculty as an assistant professor in 1997 and has since become a world leader in his field. Professor Amarasinghe's research interests span programming languages, compiler design, and high-performance computing, with a particular focus on domain-specific languages. His group has developed numerous influential languages and compilers including Halide, TACO, Simit, StreamIt, StreamJIT, PetaBricks, MILK, Cimple, and GraphIt, which deliver unprecedented performance for application domains such as image processing, stream computations, and graph analytics. He has also pioneered the application of machine learning for compiler optimizations, from Meta optimization in 2003 to the OpenTuner autotuner framework. Professor Amarasinghe's publication history reveals a consistent research trajectory toward creating specialized language and compiler solutions that address performance challenges in specific domains while hiding complexity from application developers. His recent work focuses heavily on sparse computing, tensor algebra, graph processing, and the integration of machine learning techniques into compiler technology, demonstrating his ability to identify and address emerging computational challenges. ACM Fellow (2019) As an educator, Professor Amarasinghe has developed the popular Performance Engineering of Software Systems (6.172) class with Professor Charles Leiserson and created innovative project-based courses including the Open Source Software Project Lab, the Open Source Entrepreneurship Lab, and the Bring Your Own Software Project Lab. He also serves as the faculty director of MIT Global Startup Labs, which has helped create more than 20 startups across 17 countries. His research has translated into practical applications through startups like Determina, Inc. (acquired by VMware), demonstrating the real-world impact of his academic work. Professor Amarasinghe co-led the Raw architecture project with Professor Anant Agarwal, which did pioneering work on scalable multicores. His entrepreneurial activities include founding Determina, Inc. based on computer security research from his MIT lab and co-founding Lanka Internet Services, Ltd., the first Internet Service Provider in Sri Lanka, showcasing his ability to bridge academic research with commercial applications.

Gaurav RattanView profile

Assistant Professor

• Graph Theory
• Machine Learning on Graphs
• Graph Neural Networks
• +5 more

Gaurav Rattan is an Assistant Professor in the Department of Applied Mathematics at the University of Twente's Faculty of Electrical Engineering, Mathematics and Computer Science (EEMCS), where he joined in May 2024. His research focuses on the mathematical foundations of machine learning on graphs and discrete structures, with particular emphasis on theoretical aspects of graph neural networks. University of Twente, Department of Applied Mathematics (May 2024-present) TU Darmstadt, Postdoctoral Researcher in Pascal Schweitzer's group RWTH Aachen, DFG Eigene Stelle Researcher in Martin Grohe's group Dr. Rattan completed his PhD at IMSc Chennai under V. Arvind and earned his B. Tech. from IIT Bombay, establishing a strong foundation in theoretical computer science and mathematics. His research spans graph theory, algorithms, and machine learning on graphs, with specific expertise in graph isomorphism, graph homomorphisms, and the theoretical underpinnings of graph neural networks. He applies mathematical techniques from logic and algebra to develop theory-driven approaches for graph learning systems, with practical applications in optimization, bioinformatics, and databases. Dr. Rattan's publication record reveals a consistent focus on the intersection of theoretical computer science and machine learning. His recent work explores Weisfeiler-Leman algorithms, symmetry breaking techniques, and parameterized complexity of graph problems, demonstrating how classical graph algorithms connect with modern graph learning methodologies. His research provides crucial theoretical foundations for understanding the capabilities and limitations of graph neural networks. Active in the academic community, Dr. Rattan regularly presents at conferences including the Netherlands Mathematical Congress, SIGAlgo, LOGAMS, and specialized workshops on graph learning. Recent presentations include "From Graph Homomorphisms Densities to Graph Learning" at the Graph Learning Workshop at NITMB Chicago and "Color Refinement: One Algorithm, Many Facets" at SIGAlgo 2024.

Risi KondorView profile

Associate Professor

• Machine Learning
• Graph Theory
• Quantum Mechanics
• +6 more

Risi Kondor is an Associate Professor in the Departments of Statistics and Computer Science at the University of Chicago. His research focuses on machine learning, group theory applications, and equivariant neural networks. He develops algorithms respecting geometric and physical symmetries, with contributions to graph learning, quantum mechanics modeling, and multiresolution matrix factorization. Key projects include the development of Covariant Compositional Networks (CCNs) for graph-structured data and N-body networks for molecular simulations. He has created software tools like GraphFlow, SnOB (FFT for symmetric groups), and Mondrian for high-performance computing. His work bridges algebraic methods (e.g., Fourier analysis on permutation groups) with machine learning, addressing challenges in multi-object tracking, computer vision, and materials science. Risi Kondor holds grants including a DARPA Young Faculty Award ($500K, 2016–2018) and NSF funding for non-commutative harmonic analysis in machine learning. His research emphasizes theoretical foundations and practical applications, advancing areas like equivariant architectures, multiscale analysis, and symmetry-aware machine learning systems.

Cevdet AykanatView profile

Professor

• Parallel Computing
• High-Performance Computing
• Graph and Hypergraph Partitioning
• +4 more

Prof. Cevdet Aykanat is a Professor of Computer Engineering at Bilkent University, Ankara, Turkey. He earned his BS/MS in Electrical Engineering from METU and PhD from Ohio State University as a Fulbright scholar. His research focuses on parallel computing, sparse matrix algorithms, graph partitioning, and high-performance computing for big data. He has been affiliated with Bilkent since 1989 and has held roles like Associate Provost. Education: BSc/MSc (METU Electrical Engineering), PhD (Ohio State University Electrical & Computer Engineering). Research interests include parallel scientific computing, combinatorial optimization, distributed systems, and large-scale data analysis. His work spans over 100 publications in top journals like IEEE Transactions and SIAM, with 5,500 citations and an H-index of 40. Awards: 1996 TUBITAK Investigator Award, 2007 METU Parlar Science Award. He led 6 TUBITAK projects and participated in EU-funded projects like PRACE-1IP to PRACE-6IP. Publications emphasize efficient parallel algorithms for sparse computations, graph partitioning, and distributed systems. His work addresses latency reduction, load balancing, and scalable data processing in HPC environments. Grants: Funded by TUBITAK, Intel SSD, and EU programs. Academic service includes editorial roles at IEEE Transactions on Parallel and Distributed Systems.

• Signal Processing
• Machine Learning
• Neural Networks
• +3 more

Professor Danilo P. Mandic, affiliated with Imperial College London, UK, is a leading researcher in signal processing, machine learning, and biomedical signal analysis. His work spans quaternion algebra, tensor networks, and neural networks for real-world applications. 2025: Published 11+ works on EEG/PPG analysis, quantum learning, and tensor-based LLM compression 2024: Active in interpretable transformers, graph learning for financial data, and hearable devices Research focuses on hypercomplex signal processing, graph neural networks, and medical AI applications. Recent work explores quaternion calculus for signal processing, tensor network structures for LLMs, and hearable device optimization. Key publication trends include: 2025 emphasis on quantum-aware learning, 2024 graph-based time series clustering, and 2023 foundational work on graph CNNs and matched filtering approaches. Collaborates extensively with Dongpo Xu, Sayed Pouria Talebi, Clive Cheong Took, and Tobias Reichenbach on projects involving ear-EEG, ECG enhancement, and financial sentiment analysis.

Jonathan A. KelnerView profile

Professor

• Algorithms
• Complexity Theory
• Spectral Graph Theory
• +4 more

Jonathan A. Kelner is a Professor of Applied Mathematics at the Massachusetts Institute of Technology (MIT) and a member of the Computer Science and Artificial Intelligence Laboratory (CSAIL) . His research bridges pure mathematics and algorithms, focusing on spectral graph theory, combinatorial optimization, and distributed computing.

• Computer Architecture
• Parallel Processing
• FPGA Design
• +3 more

Frank Hannig is a Professor at the University of Erlangen-Nuremberg, Germany, specializing in computer architecture and high-performance computing. His research focuses on FPGA acceleration, hardware-software co-design, neural network optimization, and embedded systems. He has collaborated extensively with co-authors such as Jürgen Teich and Oliver Reiche, producing over 200 publications since 2001. His work emphasizes domain-specific languages (DSLs) for image processing (e.g., Hipacc) and compiler optimizations for FPGAs. Key contributions include techniques for quantized neural networks on microcontrollers, CGRA toolchain evaluation, and efficient mapping of CNNs onto processor arrays. Hannig also explores energy-efficient architectures and reconfigurable computing for emerging applications like edge AI and automotive systems. Publications span conferences like ASAP, FPL, and ARC, reflecting his interdisciplinary approach to bridging algorithm design and hardware implementation. His research often addresses practical challenges in deploying machine learning models on resource-constrained devices while maintaining performance and energy efficiency.

Fredrik KjolstadView profile

Assistant Professor

• Compilers
• Programming Models
• Systems
• +5 more

Fredrik Kjolstad is an Assistant Professor of Computer Science at Stanford University. His research focuses on compilers, programming models, and systems for sparse computing, with an emphasis on separating algorithms from data representation. He leads efforts in developing compilers like TACO, Simit, and Legate Sparse to optimize sparse tensor algebra and distributed computations. Kjolstad's work spans compiler design, hardware-software co-design, and programming languages for high-performance computing. His research group aims to enable portable applications across diverse data representations and architectures. Notable projects include the TACO compiler for sparse tensor algebra, the Simit programming language for physical simulations, and the Copy-and-Patch compilation technique for fast runtime code generation. He has also contributed to distributed systems like Legate Sparse and hardware accelerators such as Onyx. Kjolstad has received prestigious awards including the NSF CAREER Award and the MIT EECS PhD Thesis Award. His publications cover topics like sparse tensor compilation, compiler optimization, and agile hardware design. He advises multiple PhD students and collaborates with researchers like Kunle Olukotun and Alex Aiken. Key areas of impact include efficient sparse data processing, compiler-driven hardware design, and scalable distributed computing frameworks. His work bridges theoretical compiler techniques with practical system implementations, aiming to simplify and accelerate complex computational tasks.

Mark IwenView profile

Professor

• Computational Harmonic Analysis
• Mathematical Data Science
• Signal Processing
• +12 more

Mark Iwen is a Full Professor with a dual appointment in the Department of Mathematics and the Department of Computational Mathematics, Science and Engineering (CMSE) at Michigan State University. His research bridges theoretical mathematics and computational science, focusing on high-dimensional data analysis, signal processing, and numerical algorithms. Research Interests: His work centers on computational harmonic analysis, mathematical data science, and algorithms for large-scale data. Key areas include sparse spectral methods, phase retrieval, tensor decomposition, compressive sensing, sublinear-time algorithms, and Johnson-Lindenstrauss embeddings. He develops provably accurate and efficient algorithms for problems in high dimensions, such as solving PDEs, function approximation, and manifold learning. Publications and Trends: His recent publications (2023–2025) emphasize sparse and low-rank tensor methods, phase retrieval, and high-dimensional PDE solvers. There is a strong focus on algorithmic efficiency, provable guarantees, and applications in imaging and data science. Many works involve collaborative efforts and include open-source code. Scientific Awards: Scholarship established in honor of Benjamin (Ben) Segal, reflecting his mentorship and academic impact Advising and Grants: He advises numerous students, including undergraduates and graduate researchers, many of whom are co-authors on publications and code repositories. While specific grants are not listed, his sustained research output and collaborative projects suggest active funding. He co-founded the One World MINDS Seminar, demonstrating leadership in the academic community. Labs and Teams: He leads a research group focused on algorithm development for high-dimensional problems, with strong ties to computational mathematics and data science. His team develops and shares code for sparse FFTs, phase retrieval, tensor methods, and PDE solvers, indicating a collaborative and open research culture.