Ismael Castillo is a Professor of Statistics at Sorbonne Université , affiliated with the Laboratoire de Probabilités, Statistique et Modélisation (LPSM) and its Statistics, Data, Algorithms team. He serves as Associate Editor for Annals of Statistics , Bernoulli , and co-Editor for Bayesian Analysis . Research Interests : Mathematical statistics with emphasis on Bayesian nonparametrics , inference in high-dimensional structures , uncertainty quantification , and applications in signal processing and life sciences . Recent Work spans deep neural networks with heavy-tailed weights , posterior and variational inference , fractional posteriors in semiparametric models , and deep Gaussian processes . His publications demonstrate expertise in multiple testing procedures , Spike and Slab priors , and nonparametric Bayesian methods . Awards : IMS Fellow , Honorary Fellow of Institut Universitaire de France , and Best Paper Prize (2021) for research on Pólya tree posterior distributions. Students : Supervised PhD candidates Paul Egels , Thibault Randrianarisoa , and co-supervised Bo Ning (FSMP postdoc) and Kweku Abraham (Hadamard postdoc). Grants : ANR BACKUP (2023-2027, coordinator) and ANR GAP (2021-2025, member).
Joseph A. Campbell is an Assistant Professor in the Department of Computer Science at Purdue University, leading the Collaborative AI for Machines and People (CAMP) Lab. He holds a Ph.D., M.S., and B.S. in Computer Science and Computer Engineering from Arizona State University. Before academia, he worked as a software engineer for five years. Prior to Purdue, he was a Postdoctoral Fellow at Carnegie Mellon University's Robotics Institute. His research focuses on explainable machine learning and robotics, particularly how agents use explanations for self-improvement and decision-making. Key areas include theory of mind in multi-agent systems, lifelong learning, and interpretable transfer learning. His work bridges robotics and AI, with applications in human-robot interaction and prosthetic control. Notable publications include advancements in reinforcement learning with language models, multi-agent collaboration frameworks, and methods for enhancing state estimation in robots. His research has been presented at top conferences like NeurIPS, EMNLP, and CoRL. Dr. Campbell maintains an active GitHub profile (joe-campbell) with repositories such as Interaction Primitives for robotics applications. His lab, CAMP, explores AI systems that collaborate effectively with humans and other machines.
Dr. Sonia Petrone is a Full Professor of Statistics at Bocconi University's Department of Decision Sciences. She earned her PhD in Statistics from Bocconi University and has held academic positions at the University of Pavia and University of Insubria before joining Bocconi. Her extensive international experience includes research visits across North America, Latin America, Europe, India, and Russia. Her research specializes in Bayesian statistics, with contributions to foundational theory, predictive modeling, Bayesian nonparametrics, and stochastic processes. She currently directs the Bocconi Summer School in Advanced Statistics and Probability and previously led the PhD program in Statistics (2011-2018). Her research portfolio demonstrates consistent focus on Bayesian nonparametric methods, predictive modeling, and applications to complex data structures. Recent work explores urn processes, time series analysis, and network modeling using innovative Bayesian approaches. Awards & Honors: IMS Medallion Lecture Award (2018) ISBA Foundational Lecture Award (2016) Fellow of International Society for Bayesian Analysis Fellow of Institute of Mathematical Statistics Fellow of European Laboratory for Intelligent Systems Fellow of Bocconi Institute of Data Science She has held editorial leadership positions as Editor of Statistical Science (2020-2022) and Bayesian Analysis (2010-2014), and served as President of the International Society for Bayesian Analysis (2014).
Aaron Smith is an Associate Professor in the Department of Mathematics and Statistics at the University of Ottawa, affiliated with the Faculty of Science. He holds a PhD from Stanford University. His research focuses on applied probability, computational statistics, Monte Carlo methods, and Markov chains, with an emphasis on advancing theoretical understanding and practical applications of these methodologies. Dr. Smith's work includes contributions to community detection algorithms, Markov chain mixing times, and synthetic health data generation. His recent publications explore topics such as nonstandard Dirichlet form representations, perturbation analysis of MCMC algorithms, and sparse Bayesian multidimensional scaling. He advises students in applied probability and has supervised postdoctoral researchers in related fields. His research interests span a wide range of topics, including stochastic processes, statistical inference, and algorithm design. He is particularly known for his analysis of convergence rates in Markov chains and the development of efficient sampling techniques for complex models. His interdisciplinary work bridges theoretical mathematics and practical computational challenges in data science and healthcare. Dr. Smith collaborates on projects involving synthetic data frameworks for privacy-preserving applications and has contributed to foundational work on mixing times and perturbation effects in stochastic systems.
Debdeep Pati is a Professor in the Department of Statistics at the University of Wisconsin-Madison, affiliated with the School of Computer, Data & Information Sciences. His research focuses on Bayesian methods, high-dimensional data analysis, machine learning, and computational statistics, with applications in health data and network analysis. He has contributed to approximate Bayesian computation, graphical models, and fair algorithms. Key research interests include Bayes theory in high dimensions, hierarchical modeling, efficient Bayesian computation, and real-time tracking algorithms. His work bridges theoretical advancements with practical applications in areas like electronic health records and nuclear physics constraints. Recent work emphasizes Wasserstein-guided nonparametric Bayes, fair clustering algorithms, and variational inference in singular models. He has developed software for covariate-dependent Gaussian graphical modeling, published in ACM Transactions on Mathematical Software . Grants: NSF proposal on Wasserstein-guided nonparametric Bayes, NIH R01/R21 grants on periodontal disease and diabetes comorbidity. Advising: No named advisees listed but actively supervising research in Bayesian computation and high-dimensional statistics. Awards: 2024 JASA reproducibility award for 'Covariate-Assisted Bayesian Graph Learning.' He is an Associate Editor for Journal of Computational and Graphical Statistics and has organized workshops at Banff International Research Station (BIRS) and the Institute for Mathematics and its Applications (IMSI).
Per Christian Hansen is a Professor at the Department of Applied Mathematics and Computer Science (DTU Compute), Technical University of Denmark (DTU), where he leads the Section for Scientific Computing. He is a VILLUM Investigator and heads the CUQI (Computational Uncertainty Quantification for Inverse Problems) research initiative, aiming to develop accessible computational platforms for uncertainty quantification in inverse problems. His expertise lies in numerical analysis, numerical linear algebra, iterative reconstruction methods, and computational inverse problems, with applications in tomography, signal analysis, and plasma physics. His research integrates theoretical analysis—such as perturbation and convergence analysis—with the development of robust, adaptive, and efficient computational methods. He has co-authored five books, over 100 scientific papers, and several widely used MATLAB software packages, including IR Tools and Regularization Tools. His recent work (2023–2025) emphasizes uncertainty quantification, Bayesian inversion, and high-dimensional tomography in fusion plasmas, reflecting a strong trend toward probabilistic and robust modeling in inverse problems. He is a SIAM Fellow (2015) for his contributions to computational methods for rank-deficient and discrete ill-posed problems and regularization techniques. His scientific leadership is evident in both theoretical advances and practical software implementations. He actively collaborates across disciplines, particularly in nuclear fusion and medical imaging, and continues to supervise PhD students and publish in top-tier journals such as Inverse Problems , SIAM Journal on Scientific Computing , and Nuclear Fusion . SIAM Fellow (2015) VILLUM Investigator He advises PhD and Master’s students in computational mathematics and inverse problems, and his research is supported by major grants, including the VILLUM Investigator award. He leads the CUQI team, which develops open-source tools for non-experts to apply uncertainty quantification in inverse problems. His lab focuses on creating modeling frameworks that bridge theory, computation, and real-world applications in materials science, imaging, and plasma diagnostics.
Jonathan Shihao Ji is an Associate Professor in the School of Computing at the University of Connecticut (UConn), leading the Intelligent Systems Lab. He holds a Ph.D. in Electrical and Computer Engineering from Duke University and previously served as an Associate Professor at Georgia State University and Director of the DoD Center of Excellence (CiARE). His research focuses on deep learning applications in computer vision, NLP, robotics, and high-performance computing, with over 50 publications in top venues like CVPR, NeurIPS, and IEEE journals. He has secured grants from NSF, NIH, DoD, and industry partners including VMware and Nvidia. His work emphasizes efficient algorithms for large-scale data processing, parameter-efficient model fine-tuning (e.g., VB-LoRA), and 3D perception benchmarks for UAVs (UAV3D). Notable contributions include sparse network optimization (Dep-L0), energy-based models (M-EBM), and robust defenses against adversarial attacks (Defense-VAE). He is a Senior Member of IEEE and has developed open-source tools like Parallel Word2Vec and WordRank. Recent projects include accelerating Llama2 models on FPGAs (LlamaF) and improving text-to-image synthesis via contrastive learning. His research spans theoretical advancements and practical applications, with industry collaborations in healthcare, robotics, and embedded systems.
Meng Li is the Noah Harding Associate Professor of Statistics at Rice University's School of Engineering. He specializes in Bayesian analysis, machine learning, and statistical theory. His research bridges methodological development and applications in biomedical sciences, materials informatics, and neuroimaging. Li holds a Ph.D. from North Carolina State University and a B.S. from Sun Yat-sen University. He has been recognized with awards including the 2020 Rice Engineering Excellence Award and the Ralph E. Powe Junior Faculty Enhancement Award. Li's research focuses on probabilistic modeling of complex data such as images, functional data, and networks. His funded projects include AI frameworks for pancreatic cancer biomarkers and Bayesian spatiotemporal modeling of marine ecosystems. He collaborates with institutions like Houston Methodist and Baylor College of Medicine on medical applications. His teaching includes advanced courses like Bayesian Statistics and Advanced Bayesian Inference. He advises over 30 students, many of whom have pursued academic and industry roles. Li serves as an associate editor for Bayesian Analysis and the new ACM Transactions on Probabilistic Machine Learning.
Malay Ghosh is a Distinguished Professor in the Department of Statistics at the University of Florida. He holds a B.A. (1962) and M.A. (1964) in Statistics from Calcutta University, and a Ph.D. (1969) in Statistics from the University of North Carolina at Chapel Hill. His research focuses on Bayesian statistics, small area estimation, and survey sampling methodologies. Ghosh has contributed extensively to statistical theory and applications, including foundational work in probability matching priors and generalized linear models for small area estimation. Research Interests : His key areas include advanced statistical modeling, methodological developments in survey sampling, and Bayesian approaches to complex data analysis. His work bridges theoretical rigor with practical applications in diverse fields requiring precise estimation techniques. Awards: Fellow, American Statistical Association Fellow, Institute of Mathematical Statistics Elected Member, International Statistical Institute Recipient of TIP (1994) and PEP (1996) Awards Editorial Roles: Editor of Sequential Analysis (since 1996), Co-Editor of Sankhya (since 2000), and Associate Editor of the American Statistician (since 2000). His editorial contributions reflect his leadership in advancing statistical discourse.
James Anderson is an Assistant Professor in the Department of Electrical Engineering at Columbia University, with affiliations to the Data Science Institute (DSI) and multiple research centers including the Computing Systems for Data-Driven Science and Foundations of Data Science. Prior to Columbia, he was a Senior Research Scientist at Caltech’s Computing + Mathematical Sciences division (2016–2019) and held a Junior Research Fellowship at the University of Oxford’s Department of Engineering Science (pre-2012). He earned his DPhil (PhD) in Engineering Science from Oxford in 2012. His research focuses on optimal/robust control theory, mathematical programming, data privacy, and cyber-physical systems, with applications in smart grids, systems biology, and power systems. Recent work emphasizes energy storage strategies, distributed control algorithms, and cybersecurity in critical infrastructure. His publications span advanced control methodologies (e.g., reinforcement learning for LQR problems), energy market dynamics, and resilient system designs. Notable contributions include frameworks for decision-focused energy storage arbitrage and defenses against false data attacks in power grids. He actively collaborates on federated learning approaches for distributed systems and has pioneered techniques for system-level synthesis in cyber-physical architectures. Anderson’s affiliations include the Data Science Institute (DSI) and specialized centers focused on data-driven science and energy systems. His work bridges theoretical control advancements with real-world applications in energy and healthcare.
Richard Nickl is a Professor of Mathematical Statistics at the University of Cambridge, affiliated with the Department of Pure Mathematics and Mathematical Statistics (DPMMS) and the Statistical Laboratory. His research focuses on high-dimensional inference, Bayesian nonparametrics, statistics for partial differential equations, and inverse problems. He has held significant grants, including an ERC Advanced Grant (2024–2029) and an EPSRC Programme Grant (2022–2027). His work bridges statistics, probability, and analysis, with contributions to theoretical foundations and computational methods in non-linear inverse problems. Key research interests include Bayesian posterior consistency, statistical inference for diffusions, and polynomial-time algorithms for high-dimensional posteriors. Notable publications include foundational monographs such as Mathematical foundations of infinite-dimensional statistical models (2016), which earned a PROSE Award, and recent advancements in Bayesian nonparametric inference for McKean-Vlasov models (2025). His group organizes workshops, such as the 2024 Statistical Aspects of Non-Linear Inverse Problems conference. Awards: 2017 PROSE Award in Mathematics. Grants: ERC Advanced Grant, EPSRC Programme Grant. Lab/Team: Research Group in Mathematical Statistics at DPMMS, focusing on inverse problems and Bayesian methodology.
Ravid Shwartz-Ziv is an Assistant Professor and Faculty Fellow at NYU's Center for Data Science, with a dual role as Senior Research Scientist at Wand AI. His research bridges theoretical foundations and practical applications in artificial intelligence, focusing on Large Language Models (LLMs), information theory, and neural network interpretability. Ph.D. in Computational Neuroscience, Hebrew University of Jerusalem (2021) B.Sc. in Computer Science and Computational Biology, Hebrew University of Jerusalem (2014) His research spans: Developing min-p sampling for LLM text generation Preventing representation collapse in Transformers Creating contamination-free LLM benchmarks like LiveBench Advancing information-theoretic frameworks for neural networks Exploring representation learning and model adaptation Recent publications demonstrate expertise in LLM efficiency, self-supervised learning, and multi-agent systems. Notable awards include the Google PhD Fellowship, Moore-Sloan Fellowship, and multiple best paper recognitions. He has led research initiatives at Intel and Google AI, focusing on neural network compression, XGBoost comparisons for tabular data, and innovative benchmarking frameworks.
Jian Peng is an Assistant Professor in the Department of Computer Science at the University of Illinois at Urbana-Champaign. His research focuses on computational biology, machine learning, and their applications to protein structure prediction, drug design, and molecular modeling. He has contributed to advancements in antibody engineering, protein-ligand docking, and generative models for biological systems. Key research areas include: Machine Learning for Molecular Modeling Protein Structure Prediction Antibody and Peptide Design Genomics and Single-Cell Analysis Structure-Based Drug Discovery His work emphasizes integrating deep learning techniques with biological datasets to address challenges in precision medicine, drug development, and systems biology. Notable achievements include developing the FastFold system to accelerate AlphaFold training and pioneering flow-based methods for antibody design. Awards include the Overton Prize (2020), recognizing contributions to computational biology. His research has been published in top journals and conferences, spanning topics from protein mutation prediction to geodesic-based immune complex modeling.
Luigi Acerbi is an Associate Professor in the Department of Computer Science at the University of Helsinki, where he leads the Machine and Human Intelligence research group. He is also an active member of the Finnish Center for Artificial Intelligence (FCAI) and ELLIS (European Laboratory for Learning and Intelligent Systems). His research focuses on probabilistic machine learning and computational neuroscience, particularly on developing efficient methods for statistical inference, Bayesian models of perception, and resource-constrained rationality. His work bridges machine learning and cognitive science, with applications in Bayesian optimization, simulation-based inference, and image completion. The recent publications highlight a strong trend toward unifying probabilistic conditioning across diverse tasks using transformer-based meta-learning frameworks like the Amortized Conditioning Engine (ACE). These works emphasize amortized inference, flexible latent variable modeling, and the integration of prior knowledge at runtime, enabling efficient and scalable Bayesian methods for complex problems. Scientific Affiliations: University of Helsinki, Department of Computer Science Finnish Center for Artificial Intelligence (FCAI) ELLIS (European Laboratory for Learning and Intelligent Systems) Education: PhD in Computational Neuroscience, Doctoral Training Centre, Edinburgh, UK Advisor: Sethu Vijayakumar and Daniel Wolpert Visiting work at Computational and Biological Learning Lab, Cambridge Postdoctoral Experience: Alex Pouget’s lab, University of Geneva, Switzerland Wei Ji Ma, New York University, USA Collaboration with the International Brain Laboratory Luigi Acerbi mentors PhD students including Daolang Huang and Nasrulloh Loka, and collaborates widely with researchers such as Samuel Kaski. He has contributed to open-source tools like PyVBMC and is involved in community initiatives such as the EurIPS conference. His work is supported by grants from the Research Council of Finland, Business Finland, and the UKRI Turing AI World-Leading Researcher Fellowship. He leads a research lab focused on amortized probabilistic inference, with ongoing projects including PriorGuide and Stacked VBMC, aiming to make Bayesian methods more practical and accessible for real-world scientific and engineering applications.
Akihiko Nishimura is an Assistant Professor in the Department of Biostatistics at the Johns Hopkins Bloomberg School of Public Health. He holds a PhD from Duke University (2017) and MS and BS degrees from Stanford University (2011 and 2010). His research focuses on Bayesian methods, statistical computing, and public health data science, with applications in precision medicine and observational health data analytics. PhD, Duke University, 2017 MS, Stanford University, 2011 BS, Stanford University, 2010 Nishimura's research centers on developing advanced statistical and computational methodologies for real-world health data. His work emphasizes Bayesian inference, large-scale computing, and software development for reproducible research. He is particularly interested in using observational health data to improve clinical decision-making and advance precision medicine. He co-leads the Bayesian Learning and Spatio-Temporal modeling group (BLAST Group) and the inHealth/OHDSI Lab , collaborating with clinicians and data scientists across institutions. His recent publications reflect a strong trend in methodological innovation in Monte Carlo methods (e.g., Hamiltonian and Zigzag samplers), scalable Bayesian inference, and applications in pharmacovigilance, diabetes management, and infectious disease modeling. The articles span disciplines including biostatistics, computational statistics, public health, and bioinformatics, demonstrating a consistent focus on high-impact, computationally intensive problems in health data science. Nishimura actively contributes to the scientific community through methodological development and open science. He develops statistical software and shares teaching materials on GitHub, emphasizing reproducibility and performant computing. His involvement in the OHDSI community enables large-scale, multi-institutional studies that would not be feasible with single-source data. His work has been recognized through publications in top-tier journals such as the Journal of the American Statistical Association , Biometrika , and JAMA Ophthalmology , and has been picked up by numerous news outlets and social media platforms, indicating broad scientific and public impact. Nishimura teaches courses on performant statistical computing and advanced Monte Carlo methods, training the next generation of data scientists in efficient algorithm and software design. He mentors students and collaborators in statistical methodology and software development, fostering a culture of rigorous, reproducible, and impactful research.
