جست‌وجوی استاد و پژوهشگر

• Data Science
• Machine Learning
• Computational Harmonic Analysis
• +۶ مورد دیگر

Pierre Vandergheynst is a Full Professor at the Swiss Federal Institute of Technology Lausanne (EPFL) in the Department of Electrical Engineering, with a courtesy appointment in Computer and Communication Sciences. He serves as EPFL’s Vice-Provost for Education since 2015 and leads the Signal Processing Laboratory 2 (LTS2). His research spans harmonic analysis, sparse approximations, mathematical data processing, and applications in signal/image processing, computer vision, machine learning, and graph-based data analysis. PhD in Mathematical Physics (1998), Université catholique de Louvain Postdoctoral Researcher at EPFL (1998-2001) Assistant Professor at EPFL (2002-2007) His research explores geometry/symmetry in high-dimensional data, redundant dictionaries for dimensionality reduction, and computational harmonic analysis on manifolds. Recent work focuses on protein structure modeling, geometric deep learning, and graph-based signal processing. Key article trends include graph neural networks for protein analysis, geometric deep learning in neuroscience, and structured knowledge priors in neural models. His 2023-2025 publications emphasize interpretable AI, long-range dependencies in graphs, and molecular representation learning. Scientific Awards: IEEE Signal Processing Magazine Best Paper Award (2023) Signal Processing Society Best Paper Award (2022) Apple ARTS Award (2007) De Boelpaepe Prize, Royal Academy of Sciences of Belgium (2009-2010) He has supervised over 30 PhD theses and contributed to foundational work in graph signal processing, compressive sensing, and geometric deep learning. His lab develops tools for data science on non-Euclidean structures, with applications in medicine, astronomy, and wireless systems.

Rex Yingمشاهده پروفایل

استادیار

• Graph Neural Networks
• Geometric Representation Learning
• Explainable AI
• +۳ مورد دیگر

Rex Ying is an Assistant Professor in the Department of Computer Science at Yale University's School of Engineering & Applied Science. He leads research in graph neural networks, geometric representation learning, and explainable AI, with applications spanning physical simulations, biology, knowledge graphs, and recommender systems. His lab actively recruits PhD students interested in geometric deep learning, graph neural networks, and trustworthy AI. Dr. Ying received his PhD in Computer Science from Stanford University under Jure Leskovec, with a thesis titled "Towards Expressive and Scalable Deep Representation Learning for Graphs." Prior to that, he graduated from Duke University in 2016 with highest distinction, majoring in Computer Science and Mathematics. His research focuses on three interconnected areas: advancing graph neural network architectures for improved expressiveness, scalability, and interpretability; innovating in geometric representation learning for data with diverse characteristics; and developing real-world applications across scientific domains. He has pioneered influential algorithms including GraphSAGE, PinSAGE, and GNNExplainer, and developed the first billion-scale graph embedding services at Pinterest as well as graph-based anomaly detection algorithms at Amazon. His recent publication trends show a strong focus on hyperbolic geometry for foundation models, non-Euclidean representation learning, and multimodal applications in computational biology. The research demonstrates increasing integration of geometric deep learning with large language models and foundation model architectures. KDD 2022 Dissertation Award 2019 Baidu Scholarship in Artificial Intelligence Dr. Ying actively serves the research community as a committee member for major conferences including AAAI, ICML, NeurIPS, ICLR, KDD, and WebConf for over seven years, and as area chair for LoG 2022. He co-leads the open-source PyTorch Geometric project and has organized numerous workshops on graph learning. His industry collaborations include Pinterest, Amazon, Facebook AI Research, DeepMind, Siemens, SLAC National Accelerator Laboratory, and Saudi Aramco. He teaches "Deep Learning for Graph-Structured Data" at Yale and mentors students in developing cutting-edge graph learning algorithms. His research lab collaborates with both academic institutions and industry partners to advance the state-of-the-art in graph representation learning, with particular emphasis on geometric deep learning and its applications to scientific discovery and real-world systems.

David Eppsteinمشاهده پروفایل

استاد

• Graph Algorithms
• Computational Geometry
• Graph Theory
• +۳ مورد دیگر

David Eppstein is a Distinguished Professor of Computer Science at the University of California, Irvine (UCI), affiliated with the Donald Bren School of Information & Computer Sciences. He holds academic leadership roles as director of the Center for Algorithms and Theory of Computation and associate director of the Center for Algorithms, Combinatorics, and Optimization. His research focuses on graph algorithms, computational geometry, discrete mathematics, and geometric graph theory. Eppstein earned a B.S. in Mathematics from Stanford University (1984) and a Ph.D. in Computer Science from Columbia University (1989). Research Interests: Graph drawing, information visualization, dynamic graph algorithms, mesh generation, optimal triangulation, K-shortest paths, subgraph isomorphism, data depth, exponential-time algorithms for NP-hard problems. Awards: ACM Fellow (2012), AAAS Fellow (2017), Distinguished Professor (2020), Best Paper Awards (2023, 2022), and SIAM recognition. Grants: Co-PI on a $1.2M NSF grant (2022) for geometric graph research, and previous NSF grants for algorithm studies (2016). His work bridges theoretical computer science and practical applications, including contributions to graph visualization, geometric algorithms, and combinatorial optimization. Notable recent achievements include resolving open questions in graph biplanarity and authoring the book Forbidden Configurations in Discrete Geometry (2018).

Rina Foygel Barberمشاهده پروفایل

استاد

• Statistics
• Statistical Inference
• Distribution-Free Inference
• +۷ مورد دیگر

Rina Foygel Barber is the Louis Block Professor in the Department of Statistics at the University of Chicago, where she also serves as Co-chair of the Committee on Community, Diversity, and Inclusion (CCDI) and is a member of the Committee on Computational and Applied Mathematics (CCAM). Her educational background includes: PhD in Statistics, University of Chicago (2012), advised by Mathias Drton and Nati Srebro MS in Mathematics, University of Chicago (2009) ScB in Mathematics, Brown University (2005) NSF postdoctoral fellow, Stanford University Department of Statistics (2012-13), supervised by Emmanuel Candès Professor Barber's research focuses on the theoretical foundations of statistical problems in estimation, prediction, and inference, particularly in high-dimensional settings where classical methods may not be reliable. She specializes in distribution-free inference methods such as conformal prediction, multiple testing methods, algorithmic stability, and shape-constrained inference. Her work also extends to modeling and optimization problems in medical imaging reconstruction. Her recent publications demonstrate a strong focus on distribution-free inference, with particular emphasis on conformal prediction, false discovery rate control, and algorithmic stability. Her work bridges theoretical statistics with practical applications, especially in the medical imaging domain. Professor Barber has received numerous prestigious awards: Elected to National Academy of Sciences (2025) MacArthur Fellowship (2023) IMS Fellow (2023) COPSS Presidents' Award (2020) Peter Gavin Hall Early Career Prize (2020) She actively mentors students and collaborators, with many co-authored publications across statistics, machine learning, and medical imaging. Her research has been supported by significant grants that enable her work on theoretical foundations of statistical inference and practical applications in medical imaging. Professor Barber also co-organizes the International Seminar on Selective Inference. Her research group focuses on developing and analyzing estimation, inference, and optimization tools for structured high-dimensional data problems. They work on false discovery rate control, distribution-free inference, and applications in medical imaging reconstruction.

• Self-organized criticality
• Neuronal network dynamics
• Excitation/Inhibition balance
• +۵ مورد دیگر

Anna Levina is an Assistant Professor for Computational Neuroscience at the University of Tübingen , affiliated with the Department of Computer Science under the Faculty of Science. Her research focuses on the self-organization of neuronal activity, critical dynamics in neural networks, and the excitation/inhibition balance in cortical circuits. Current positions: Assistant Professor (since 2018), Group Leader (2017-2018), Equality Officer (Computer Science) Previous roles: IST Fellow (2015-2017), Associated Researcher (2011-2015), Postdoc/PI (2011-2015), Postdoc (2008-2011) Her research integrates mathematical modeling , statistical physics , and computational neuroscience to study criticality phenomena, neural avalanches, and adaptive network dynamics. Key interests include: Self-organized criticality in neural systems Excitation/Inhibition balance mechanisms Network topology and dynamics Timescale analysis in neural processing Stochastic modeling of neural activity Recent publications reveal trends in understanding critical dynamics across biological and artificial networks, with applications to memory systems, sensorimotor integration, and disease modeling. She has received recognition as an IST Fellow .

Melanie Weberمشاهده پروفایل

استادیار

• Geometric Machine Learning
• Optimization on Manifolds
• Non-Euclidean Machine Learning
• +۳ مورد دیگر

Melanie Weber is an Assistant Professor of Applied Mathematics and Computer Science at Harvard University's John A. Paulson School of Engineering and Applied Sciences (SEAS), leading the Geometric Machine Learning Group. Her research focuses on leveraging geometric structures in data for designing efficient machine learning and optimization algorithms with theoretical guarantees. She holds a PhD from Princeton University (2021) and has held fellowships at the Mathematical Institute of Oxford, Brasenose College, and the Simons Institute. Her work bridges geometry, optimization, and machine learning, with funding from NSF, Sloan Foundation, and Harvard initiatives. Education : PhD in Applied Mathematics, Princeton University (2021) BSc/MSc in Mathematics and Physics, University of Leipzig (2016) Research Interests : Dr. Weber's research integrates geometric principles into machine learning and optimization, focusing on non-Euclidean spaces, graph structures, and manifold-based methods. Key areas include optimization on Riemannian manifolds, curvature-based analysis (e.g., Ricci curvature), and developing algorithms resilient to data geometry challenges like over-smoothing in graph neural networks. Her work emphasizes theoretical foundations while addressing practical scalability in high-dimensional data. Awards & Recognition : 2024 Sloan Research Fellowship 2023 Leslie Fox Prize in Numerical Analysis 2023 NSF Grant for Geometric Optimization Grants & Funding : Supported by National Science Foundation (NSF), Alfred P. Sloan Foundation, Aramont Foundation, Harvard Dean’s Fund, and Harvard Data Science Initiative. Labs & Collaborations : Leads the Geometric Machine Learning Group at SEAS, collaborating with institutions like MIT, Max Planck Institute, and industry labs (Facebook, Google, Microsoft). Active in organizing workshops on geometric methods and curvature analysis.

Pascal Mettesمشاهده پروفایل

استادیار

• Hyperbolic Deep Learning
• Computer Vision
• Vision-Language Models
• +۳ مورد دیگر

Pascal Mettes is a tenured Assistant Professor at the University of Amsterdam within the Informatics Institute, specializing in Artificial Intelligence. He leads groundbreaking research in hyperbolic deep learning, a field he has significantly advanced through theoretical developments and practical applications in computer vision and multimodal learning. His research focuses on three primary domains: hyperbolic vision-language models that address the hierarchical nature of language-vision relationships; hierarchical deep learning using hyperbolic embeddings that naturally accommodate exponential growth patterns; and robust deep learning in hyperbolic space that improves out-of-distribution detection and network resilience. Mettes has established himself as a leading figure in this emerging field through numerous publications at top-tier conferences including CVPR, ICCV, ICML, NeurIPS, and ICLR. His recent work demonstrates how hyperbolic geometry provides natural solutions to fundamental limitations in modern deep learning, particularly regarding hierarchical data structures that cannot be adequately represented in Euclidean space. The publication trends show increasing impact and recognition in the computer vision and machine learning communities, with multiple papers receiving oral presentations and best paper nominations. Best paper nomination ESWC25 for 'Designing Hierarchies for Optimal Hyperbolic Embedding' Finalist MM 2023 Best Open-Source Software Competition (for HypLL) Multiple reviewer awards across major conferences including CVPR, ICLR, ECCV, ICML, and NeurIPS MM 2016 Best Doctoral Student Award TRECVID 2015 Winner Multimedia Event Detection Benchmark Mettes actively mentors eight PhD students working on hyperbolic learning and related topics, while also securing significant research funding including ELLIs PhD Award, NWO ClickNL, Google Perception Academic Funding, and Data Science Centre PhD Grants. He serves in prominent academic roles as Program Chair for International Conference on Multimedia Retrieval 2026 and has organized multiple workshops on hyperbolic learning at major conferences. His leadership in establishing hyperbolic deep learning as a recognized research direction is evident through his survey paper in IJCV 2024 and the development of the HypLL library for hyperbolic learning.

• Theoretical Computer Science
• Error-Correcting Codes
• Approximation Algorithms
• +۵ مورد دیگر

Venkatesan Guruswami is a Chancellor's Professor in the Department of EECS and a Senior Scientist at the Simons Institute for the Theory of Computing at UC Berkeley . He also holds a Professor position in the Department of Mathematics . His academic journey began with a B.Tech in Computer Science from the Indian Institute of Technology, Madras (1997) , followed by a Ph.D. in Computer Science from the Massachusetts Institute of Technology (2001) . After a Miller Research Fellowship at UC Berkeley (2001–02), he held faculty roles at the University of Washington and Carnegie Mellon University before returning to UC Berkeley in January 2022. Education : B.Tech, IIT Madras (1997) Ph.D., MIT (2001) Professional Affiliations : Chancellor's Professor, UC Berkeley (EECS) Senior Scientist & Interim Director, Simons Institute Professor, UC Berkeley (Mathematics) Guruswami's research spans multiple domains within Theoretical Computer Science , focusing on Error-Correcting Codes , Approximation Algorithms , Randomness in Computing , Probabilistically Checkable Proofs , and Computational Complexity . His groundbreaking work in List Decoding has enabled codes with minimal redundancy for correcting worst-case errors, while recent advancements include Polar Codes , Deletion-Correcting Codes , and Constraint Satisfaction Problems . He has also contributed to Quantum Coding Theory , Locally Recoverable Codes , and Approximation Hardness in various computational contexts. His publications reflect a deep engagement with interdisciplinary topics. Key trends include: Quantum Information Theory : Quantum LDPC codes, transversal gates, and quantum storage. Algebraic Coding : Reed-Solomon codes, AG codes, and polynomial-based constructions. Computational Complexity : Hardness of approximation, CSPs, and parameterized intractability. Data Transmission : Polar codes, deletion channels, and feedback mechanisms. Algorithmic Techniques : Spectral methods, semirandom models, and Lasserre hierarchy applications. Guruswami has received numerous accolades, including the Simons Investigator Award , Presburger Award , Packard Fellowship , Sloan Research Fellowship , ACM Doctoral Dissertation Award , and the IEEE Information Theory Society Paper Award . He is an ACM Fellow (2017) and IEEE Fellow (2019) , with recent honors like the Guggenheim Fellowship (2023) and AMS Fellow (2023) . As an advisor, he has mentored over 25 PhD and postdoctoral researchers , including Atri Rudra , Prasad Raghavendra , and Peter Manohar , whose work has won awards like the Edmund M. Clarke Doctoral Dissertation Award and CRA Outstanding Undergraduate Researcher Award . His research is supported by grants from the National Science Foundation , Packard Foundation , and Sloan Foundation . He also serves as Editor-in-Chief of the Journal of the ACM and holds leadership roles in IEEE and arXiv moderation. Guruswami is actively involved in Simons Institute programs and co-organized workshops on Coded Computation and Information Theory . His work bridges theoretical advancements with practical applications in Cloud Storage , Quantum Computing , and Group Testing , including pandemic-era contributions like AC-DC: Amplification Curve Diagnostics for SARS-CoV-2 .

David Gabaiمشاهده پروفایل

استاد

• Topology
• Hyperbolic Geometry
• 3-Manifold Topology
• +۳ مورد دیگر

David Gabai is the Hughes-Rogers Professor of Mathematics at Princeton University's Department of Mathematics. His research focuses on topology, hyperbolic geometry, and low-dimensional topology, particularly in the study of 3-manifolds and knot theory. He has collaborated extensively with leading researchers in the field, addressing foundational questions in geometric topology and hyperbolic structures. His work explores topics such as Heegaard splittings, ending laminations, hyperbolic volume minimization, and the classification of exceptional 3-manifolds. Recent contributions include advancements in shrinkwrapping techniques, knot covering problems, and the interplay between geometric and topological properties of manifolds. While no specific educational background is detailed in the provided text, his affiliations and publications indicate a deep engagement with geometric topology. Awards and grants are not explicitly listed here. He has advised no listed students in the given data, though his collaborations suggest mentorship in research groups. His research often intersects with geometric structures, manifold decompositions, and topological dynamics, reflecting a commitment to advancing foundational questions in topology and geometry.

• Functional data analysis
• Non-Euclidean data analysis
• High-dimensional data analysis
• +۳ مورد دیگر

Associate Professor LIN Zhenhua serves as a Presidential Young Professor in the Department of Statistics and Data Science at the National University of Singapore (NUS), with additional affiliation at the Institute of Data Science since 2021. His research develops cutting-edge statistical methodologies for complex data structures across multiple domains. Dr. LIN completed his Ph.D. at the University of Toronto in 2017 under Fang Yao's supervision, following M.Sc. degrees from Simon Fraser University (2013, 2010) and a B.Sc. from Fudan University (2008). Ph.D., University of Toronto, 2017 (Advisor: Fang Yao) M.Sc., Simon Fraser University, 2013, 2010 B.Sc., Fudan University, 2008 His research program spans functional data analysis (developing techniques for curves and surfaces), non-Euclidean data analysis (statistical methods on manifolds), high-dimensional statistics (p > n problems), and constrained statistical modeling. LIN's work bridges theoretical statistics with practical applications through rigorous mathematical frameworks and computational implementations. Recent publications reveal strong emphasis on bootstrap methods for high-dimensional inference, Riemannian geometry approaches for manifold-valued data, and innovative functional data techniques. His research shows consistent output with multiple 2025 publications in top journals including Biometrika, Bernoulli, and Journal of the American Statistical Association. Professional Recognition Presidential Young Professor, NUS (2019-present) Associate Editor, Bernoulli (2022-2024) Associate Editor, Statistics (2023-present) Young Researchers Committee, Bernoulli Society (2020-2024) Professor LIN actively mentors graduate students as evidenced by numerous collaborative publications with trainees. He teaches advanced courses including ST5215 Advanced Statistical Theory, DSA4211 High-dimensional Statistical Analysis, and ST5223 Statistical Models across multiple academic years. His research group develops specialized software packages including hdanova, matrix-manifold, synfd, mcfda, and iRFDA, making advanced statistical methods accessible to practitioners.

• Artificial Intelligence
• Optimization
• Machine Learning
• +۷ مورد دیگر

Sebastian Pokutta is a Professor at Technische Universität Berlin, Vice President at the Zuse Institute Berlin (ZIB), and Chair of the Cluster of Excellence MATH+ and MODAL. His research lies at the intersection of Artificial Intelligence, Optimization, and Machine Learning, with applications in sustainability, quantum computing, and mathematical discovery. Research Interests: Development of novel optimization algorithms, particularly Frank-Wolfe and Conditional Gradient methods. Integration of machine learning with decision-making and combinatorial optimization. AI for Science (AI4Science), including applications in quantum mechanics and ecology. AI and creativity, human-AI co-creativity, and social science modeling using multi-agent LLMs. His recent publications (2025) demonstrate a strong focus on scalable optimization, interpretability, and algorithmic foundations. The work spans theoretical advances in convergence analysis, practical implementations in Julia (FrankWolfe.jl), and real-world deployments in biomass estimation and quantum certification. Scientific Awards: Gödel Prize (2023) STOC Test of Time Award (2022) Science Prize of the Association for Pediatric Orthopedics (2025) Google Research Awards (2021, 2020) NSF CAREER Award (2015) He advises a vibrant research group, with former students and postdocs securing faculty positions at institutions like Inria, Carlos III University, and James Madison University. His group has received funding from Google, DFG, and Math+, and he leads major collaborative efforts such as the Thematic Einstein Semester on Mathematical Optimization for Machine Learning. Labs and Teams: Interactive Optimization and Learning Lab at TU Berlin and ZIB. Leadership in MODAL and MATH+ research clusters, fostering interdisciplinary collaboration in mathematical optimization and AI.

Thomas E. Bittnerمشاهده پروفایل

دانشیار

• Formal Ontology
• Applied Ontology
• Ontology of Space and Time
• +۳ مورد دیگر

Thomas E. Bittner serves as an Associate Professor in the Department of Philosophy at the University at Buffalo, specializing in formal and applied ontology with particular focus on spatial and temporal reasoning. His academic work bridges philosophy, computer science, and geographic information systems. His research interests encompass Formal and Applied Ontology , Ontology of Space, Time and Spatial Entities , Spatio-Temporal Reasoning , and Bio-Medical Ontology . He investigates fundamental questions about spatial relations, vagueness in geographic boundaries, and quantum-inspired models for geographic information systems, developing rigorous formal frameworks applicable to both philosophical and computational domains. His publication record demonstrates consistent engagement with cutting-edge topics in ontology, particularly exploring quantum geography concepts and computational implementations of formal ontologies. Key themes across his work include the representation of spatial vagueness, quantum-inspired geographic modeling, and verifiable computational ontology systems. Bittner holds a PhD from the Technical University of Vienna and teaches courses including Logic and Spatial Ontology. His contact information lists office location at 109 Park Hall on UB's North Campus with phone (716) 645-5149.

Rekha R. Thomasمشاهده پروفایل

استاد

• Optimization
• Applied Algebraic Geometry
• Computer Vision
• +۳ مورد دیگر

Rekha R. Thomas is a Professor of Mathematics and Undergraduate Program Director at the University of Washington. She holds a Ph.D. in Operations Research from Cornell University (1994), with postdoctoral experience at Yale University and the Konrad-Zuse-Zentrum in Berlin. Her research focuses on optimization, applied algebraic geometry, and computer vision, with contributions to semidefinite programming, graphical designs, and geometric algorithms. She has held distinguished positions such as the Robert R. and Elaine F. Phelps Professorship (2008–2012) and the Robert B. Warfield Jr. Faculty Fellowship (2017–2020). Her work bridges theory and application, addressing challenges in computer vision, combinatorial optimization, and algebraic geometry. Notable contributions include advancements in multiview geometry, kernel learning, and the geometric analysis of rank-deficient matrices. She actively collaborates across disciplines, publishing extensively and supervising numerous graduate students and postdocs. Rekha also engages in academic leadership, mentoring students, and participating in international conferences. Her research has been recognized through invited talks at major events like the International Congress of Mathematicians (2018) and SIAM Annual Meetings. She continues to explore the intersections of algebraic geometry, optimization, and computational methods.

Ziming Zhangمشاهده پروفایل

استادیار

• Computer Vision
• Machine Learning
• Deep Learning
• +۹ مورد دیگر

Ziming Zhang is an Assistant Professor in the Department of Electrical and Computer Engineering at Worcester Polytechnic Institute (WPI) , with additional affiliations in Data Science and Robotics Engineering. He previously held research roles at Mitsubishi Electric Research Laboratories (MERL) and Boston University. PhD in Computing (2013) from Oxford Brookes University , UK MS in Computing Science (2010) from Simon Fraser University , CA BS in Computer Science and Technology (2005) from Northeastern University , China Research interests span computer vision , machine learning , and their applications in point cloud processing , medical imaging , autonomous driving , and IoT . He leads the Vision, Intelligence, and System Laboratory (VISLab) at WPI. Recent publications focus on 3D reconstruction , hyperbolic learning , and robust classifiers . Awards include the R&D100 Award 2018 and NSF funding for data-efficient deep learning. PhD Students: Yecheng Lyu (co-supervised), Guojun Wu (co-supervised), Hangrui Zhang, Xuechu Yu Master's Students: Yun Yue, Yuping Shao Visiting Scholars: Fangzhou Lin His lab partners with industry and academic institutions, focusing on autonomous systems , robotics , and scientific imaging projects.

Tao Zouمشاهده پروفایل

دانشیار

• Covariance regression modeling
• Network data modeling
• Financial statistics
• +۳ مورد دیگر

Tao Zou is an Associate Professor at the Research School of Finance, Actuarial Studies and Statistics, Australian National University. His research spans covariance regression modeling, network data analysis, and applications in financial and environmental statistics. He earned a Ph.D. in Statistics in 2016. Ph.D. in Statistics, 2016 Dr. Zou’s work pioneers covariance regression, where covariances are modeled as functions of covariiates. Key contributions include robust estimation techniques, spatio-temporal modeling, missing data imputation via semi-supervised learning, and distributed data aggregation. His methods address challenges in high-dimensional and non-Euclidean data analysis. Recent publications (2025–2023) explore quasi-score matching for spatial autoregressive models, regularization in network regression, functional principal component analysis for complex data, and environmental applications like PM2.5 pollution studies. These works emphasize robustness, scalability, and interdisciplinary relevance in economics, finance, and environmental science. Dr. Zou collaborates on projects like the 2023 Data Analysis App to Empower Assessment of Immunogenicity of Biologics (Co-Investigator). While his student supervision list isn’t explicitly provided, his methodological advancements influence big data and spatial statistics. He contributes to open-access software and continues expanding covariance regression for non-normal and functional data.