Harri Lähdesmäki is an Associate Professor (tenured) at the Department of Computer Science, Aalto University, where he leads the Computational Systems Biology research group. His work focuses on probabilistic machine learning and deep generative models with applications in biomedicine and molecular biology. Key Research Interests: Probabilistic machine learning, deep generative models, computational biology, bioinformatics, longitudinal data modeling Contact: harri.lahdesmaki@aalto.fi | Konemiehentie 2, 02150 Espoo, Finland His recent publications highlight advancements in: Gaussian process priors for scalable deep generative models Single-cell analysis of immune repertoires in leukemia and diabetes Probabilistic deconvolution methods for RNA-seq data Epigenetic analysis using hidden Markov and mixed models Transformer-based survival prediction and missing data handling Harri’s work integrates mechanistic modeling with Bayesian inference, particularly applied to immunology, cancer biology, and early disease prediction.
Aapo Hyvärinen is a Professor of Computer Science at the University of Helsinki , affiliated with the Helsinki Institute for Information Technology and the Helsinki Probabilistic Machine Learning Lab . He previously held the position of Professor of Machine Learning at the Gatsby Computational Neuroscience Unit, University College London (2016-2019). Education : Undergraduate Mathematics at University of Helsinki, Vienna, and Paris; Ph.D. in Information Science from Helsinki University of Technology (1997) His research focuses on machine learning and computational neuroscience , particularly: Independent Component Analysis (ICA) Natural Image Statistics Causal Representation Learning Neural Signal Processing Applications to brain imaging (MEG, CryoEM) Recent publications emphasize causal discovery , identifiable machine learning , and nonlinear ICA . Key projects include: VETURI (AI for health) DIGIMIND (AI in mental health) CIFAR grants (2022-2025) Scientific awards : Highly Cited Researcher (2010) He serves as Action Editor for the Journal of Machine Learning Research and Neural Computation , and has held Area Chair roles at NeurIPS, ICML, ICLR, AISTATS, and UAI conferences. His work bridges theoretical machine learning with neuroscience and philosophical implications of artificial intelligence .
Florian Schäfer is an Assistant Professor at the School of Computational Science and Engineering at Georgia Tech. His research spans numerical computation, statistical inference, and competitive games, with applications in materials science, turbulence modeling, computer graphics, and computational geometry. He will join the Courant Institute at NYU in September 2025. PhD in Applied and Computational Mathematics, Caltech Bachelor’s and Master’s in Mathematics, University of Bonn His work focuses on information geometric mechanics to design structure-preserving numerical methods for continuum mechanics. This includes: State-of-the-art solvers for elliptic PDEs via Gaussian elimination and conditional independence Efficient multi-agent optimization algorithms Information geometric regularization for compressible fluid dynamics Enabling the first compressible fluid simulation exceeding 100 trillion grid cells His recent research trends integrate: Machine learning for materials science (e.g., active learning, Bayesian approaches) Stochastic modeling of microstructures and phase-field problems Neural operators for super-resolution fluid dynamics Generative models for polycrystalline material datasets Optimal transport and diffusion models for conditional density transformations High-performance computing at extreme scales Florian collaborates with researchers including Houman Owhadi, Jessie Liu, Spencer Bryngelson, Tamer Zaki, and Ali Mani. He actively presents at conferences like SIAM and UCLA seminars, and is recruiting PhD students for work at the Courant Institute starting 2025.
Matthew B. Blaschko is a Professor in the Department of Electrical Engineering at KU Leuven, Belgium. He serves as director of the KU Leuven ELLIS unit and is a fellow in the ELLIS Health program. He is a Core PI in the Flanders AI Research Program, working as a workpackage lead for Decision Support Systems and Medical Imaging. Blaschko is also a member of the KU Leuven Institute for Artificial Intelligence and one of the leaders of the working group on Machine Learning and Data Science. Professor Blaschko received his B.S. from Columbia University, M.S. from the University of Massachusetts Amherst, and Dr. rer. nat. from Technische Universität Berlin (awarded for work at Max Planck Institutes Tübingen). He was a Newton International Fellow at the University of Oxford and received his Habilitation (HDR) from École Normale Supérieure de Cachan. Prior to joining KU Leuven, he was a Permanent Research Scientist in the INRIA Saclay Research Center and a Faculty Member at Ecole Centrale Paris. His research focuses on machine learning techniques applied to visual data, with particular emphasis on calibration in deep learning, medical image analysis, and federated learning. Blaschko's work bridges theoretical foundations with practical applications, as evidenced by technology developed in his research being incorporated into MONA, software for ophthalmic image analysis. His research group has made significant contributions to the fields of model calibration, uncertainty estimation, and medical imaging analysis, with recent publications showing strong trends toward improving reliability of AI systems in medical contexts and advancing theoretical understanding of calibration metrics. Professor Blaschko has been recognized with several awards including the Université Paris-Saclay STIC Doctoral School Best Scientific Contribution Award, Best Paper Award at CVPR 2008, Main Award at DAGM 2008, and Best Student Paper Award at ECCV 2008. Professor Blaschko has supervised numerous PhD and Master's students, with current and former students including Deniz Soysal, Claire Marchal, Dongli Xu, Sebastian Gruber, Jiameng Li, Marco Mezzina, and many others working on diverse topics from Alzheimer's disease analysis to surgical phase recognition. His research has been supported by various funding sources including the Flanders AI Research Program. He has co-organized several influential workshops including the "Another Brick in the AI Wall: Building Practical Solutions from Theoretical Foundations" at CVPR 2025, Commands 4 Autonomous Vehicles workshop at ECCV 2020, and the Learning from Limited Labeled Data workshop series at NIPS 2017 and ICLR 2019. His laboratory focuses on machine learning for medical image analysis, with applications in ophthalmology, neurology, and surgical robotics. The group maintains active collaborations with medical institutions and participates in international challenges such as the KNee OsteoArthritis Prediction (KNOAP2020) challenge.
Michael U. Gutmann is a Senior Lecturer in Machine Learning at the School of Informatics, University of Edinburgh, and a member of the Institute for Adaptive and Neural Computation. His research lies at the intersection of machine learning, statistics, and scientific applications, with a focus on developing inference methods for complex and implicit models. Education: PhD in Computational Neuroscience, University of Tokyo MSc in Engineering and Applied Mathematics, Swiss Federal Institute of Technology (ETH) Zurich MSc, Ecole Centrale Paris His primary research interests include Bayesian inference, likelihood-free inference, optimal experimental design, unsupervised learning, and applications in computational biology and neuroscience. He is best known for introducing Noise-Contrastive Estimation (NCE), a foundational technique for training unnormalized statistical models. His recent work spans variational inference, density ratio estimation, flow models for missing data, and AI-driven experimental design in behavioral and biological sciences. His publications, including in NeurIPS , ICML , JMLR , and eLife , demonstrate a strong emphasis on methodological innovation for scientific discovery. He has contributed to open-source tools such as ELFI (Engine for Likelihood-Free Inference) and developed practical implementations of robust inference algorithms. Scientific Awards: No specific awards listed in the provided texts. Michael Gutmann actively supervises students and collaborates with leading researchers in machine learning and computational biology. He has secured research funding from EPSRC and BBSRC for projects in generative modeling and infectious disease epidemiology. He teaches advanced courses such as Probabilistic Modelling and Reasoning and Data Mining, reflecting his deep engagement with both theoretical and applied aspects of machine learning. Labs and Research Groups: Institute for Adaptive and Neural Computation (ANC), University of Edinburgh Former affiliations with Department of Mathematics and Statistics and Department of Computer Science at the University of Helsinki and Aalto University
Dr. Hong Ge serves as a Senior Research Fellow in the Department of Engineering at the University of Cambridge and holds a Fellowship at Darwin College. He maintains a dual affiliation with The Alan Turing Institute where his research centers on probabilistic programming. His work develops foundational methodologies for machine learning and intelligence with emphasis on Bayesian inference and decision-making under uncertainty. His academic background includes: MSc at the University of Edinburgh supervised by Chris Williams PhD at the University of Cambridge supervised by Zoubin Ghahramani Dr. Ge specializes in Bayesian nonparametrics, probabilistic programming, and neural networks. His research advances computational frameworks for uncertainty quantification and large-scale probabilistic modeling, with applications spanning decision theory and Gaussian processes. He bridges theoretical machine learning with practical implementations through open-source software development. He has mentored 12 researchers including current advisees Neel Alex (co-supervised with David Krueger) and Wenlin Chen (co-supervised with Miguel Hernández-Lobato and Bernhard Schölkopf), and former members such as Alexander Terenin (now at Cornell University) and Yongchao Huang (now Lecturer at Aberdeen University). His group contributed to UK government pandemic response through the Turing-RSS Health Data Lab. Dr. Ge leads the Turing.jl team developing probabilistic programming tools including AdvancedHMC.jl for Hamiltonian Monte Carlo, NestedSamplers.jl, and AbstractGPs.jl for Gaussian processes. He co-organized the CAPP Workshop on Probabilistic Programming and ACMLL Workshop on Machine Learning Languages.
Dr. Vincent Fortuin is a tenure-track Assistant Professor at the Technical University of Munich (TUM) and a research group leader at Helmholtz AI in Munich. He leads the Efficient Learning and Probabilistic Inference for Science (ELPIS) group and holds multiple prestigious fellowships including the Branco Weiss Fellowship. His academic affiliations include the TUM School of Computation, Information and Technology, the Konrad Zuse School of Excellence in Reliable AI, and the Munich Center for Machine Learning. Dr. Fortuin earned his BSc in Molecular Life Sciences from the University of Hamburg (2012-2015), followed by an MSc in Computational Biology and Bioinformatics from ETH Zürich (2015-2017), where he received the ETH Excellence Scholarship and the Willi Studer Prize. He completed his PhD in Machine Learning at ETH Zürich (2017-2021) under the supervision of Gunnar Rätsch and Andreas Krause, supported by a Swiss Data Science Center PhD Fellowship. Prior to joining TUM, he was a Research Fellow at St. John's College, University of Cambridge (2022-2023). His research focuses on the intersection of Bayesian statistics and deep learning, specifically developing methods for more robust, data-efficient AI systems with reliable uncertainty estimates. His work addresses critical limitations in standard deep learning approaches, particularly their tendency to be overconfident in predictions and require large datasets for training. He investigates better priors and more efficient inference techniques for Bayesian deep learning, deep generative modeling, meta-learning, and PAC-Bayesian theory, with applications in scientific and biomedical domains. Dr. Fortuin's recent publications demonstrate a consistent focus on improving uncertainty quantification in deep learning systems, with increasing emphasis on practical applications in scientific contexts. His work spans from theoretical foundations of Bayesian deep learning to practical implementations in protein design, materials science, and medical applications. A notable trend is his exploration of how to make Bayesian methods more scalable and applicable to modern large-scale AI systems while maintaining theoretical guarantees. Branco Weiss Fellowship (2023) St John's College Research Fellowship (2022) Swiss National Science Foundation Postdoc.Mobility Fellowship (2022) Swiss Data Science Center PhD Fellowship (2018) ETH Excellence Scholarship (2015) Willi Studer Award (2018) Dr. Fortuin actively supervises PhD and Master's students through his ELPIS research group at Helmholtz AI. He serves as a regular reviewer and area chair for major machine learning conferences and is an action editor for TMLR. He co-organizes the Symposium on Advances in Approximate Bayesian Inference (AABI) and the ICBINB initiative, demonstrating his commitment to advancing the field through community building. His research group receives funding from multiple sources including Helmholtz AI, the Branco Weiss Fellowship, and collaborations with international institutions. Dr. Fortuin leads the Efficient Learning and Probabilistic Inference for Science (ELPIS) group at Helmholtz AI, which focuses on fundamental machine learning research motivated by real-world scientific problems. The group collaborates extensively with researchers across Helmholtz centers and international institutions, particularly in biomedical applications where reliable uncertainty estimates are crucial.
Scientia Professor Robert Kohn is a distinguished academic at the University of New South Wales, holding a position in the School of Economics within the UNSW Business School. With a career spanning several decades, Professor Kohn has established himself as a leading expert in statistical methodology and econometric modeling. His research has significantly contributed to Bayesian statistics and computational methods for complex data analysis. Professor Kohn's research focuses on advanced statistical methodologies including Bayesian methodology, variable selection and model averaging, nonparametric regression models, time series modeling, multivariate Gaussian and non-Gaussian regression, and Markov chain Monte Carlo simulation algorithms. His work bridges theoretical statistics with practical applications across economics, finance, and cognitive science. His research demonstrates a consistent trajectory toward developing more efficient computational methods for complex statistical models, with recent work emphasizing variational Bayesian methods, particle filtering techniques, and applications to time series analysis. Analysis of his recent publications (2022-2025) reveals a strong focus on advancing computational statistical methods, particularly in Bayesian inference for complex models. His work shows increasing integration of machine learning techniques with traditional statistical methods, especially in handling high-dimensional data and complex time series structures. Professor Kohn has made significant contributions to variational inference methods, particle-based computational techniques, and applications to financial time series and cognitive modeling. Professor Kohn has maintained an exceptionally productive research career with continuous publication output since the 1970s, demonstrating remarkable longevity and adaptability in his research focus as statistical methodologies have evolved. His work shows strong international collaboration, particularly with researchers in Australia, the United States, and Europe, reflecting his standing in the global statistical community.
Markus Heinonen is an Academy Research Fellow at Aalto University's Department of Computer Science within the School of Science. His academic position is tied to Harri Lähdesmäki's Professorship, focusing on probabilistic machine learning. He holds a Doctoral degree in Engineering and Technology from the University of Helsinki (2013). His research integrates probabilistic modeling , deep learning , and differential equations , with applications in computational biology, drug discovery, and biophysics. Key themes include Gaussian processes, Bayesian inference, generative models, and their use in understanding complex biological systems like immune cell behavior (e.g., T cell receptor analysis in aplastic anemia) and molecular design. He leads major projects such as the Deep Learning with Differential Equations initiative (2020–2025), exploring continuous-time models and physics-informed neural networks. His work bridges theory and application, evidenced by collaborations in diffusion models , optimal transport , and single-cell analysis . Publications span over 65 peer-reviewed outputs, with recent emphases on robust neural network training, multi-target molecular prediction, and interpretable drug design frameworks. His research contributes to UN Sustainable Development Goal 3 (Good Health) through advancements in disease modeling and therapeutic development. He has undertaken visiting research roles at the University of California, San Francisco (2017) and Telecom ParisTech (2013–2014). Media highlights include recognition for work on TCR-epitope prediction and AI-driven enzyme engineering.
Professor Carl Edward Rasmussen is affiliated with the University of Cambridge , where he focuses on Machine Learning , Probabilistic Inference , Decision Making , and Reasoning Under Uncertainty . His work bridges theoretical advancements with practical applications in robotics, control systems, and computational biology. Academic Affiliation: University of Cambridge Academic Role: Professor His research emphasizes scalable Gaussian process methods, Bayesian system identification, and reinforcement learning. Recent projects include transfer learning for antibacterial discovery , graph neural processes for molecular functions , and efficient variational inference techniques . Key themes in his publications highlight uncertainty quantification , model generalization , and nonparametric approaches . Notable Scientific Contributions Advancements in sparse Gaussian process hyperparameter estimation Framework for Bayesian system identification in dynamic systems Hybrid models combining transformers and Gaussian processes
Quoc Thong Le Gia is an Associate Professor in the School of Mathematics & Statistics at the University of New South Wales (UNSW), Sydney. He holds a PhD in Mathematics from Texas A&M University (2003), an MS in Mathematics from Texas A&M University (2000), and a BSc in Mathematics and Computer Science from UNSW (1998). His research focuses on Numerical Analysis , Approximation Theory , Partial Differential Equations , and Stochastic Processes , with particular expertise in problems on spherical domains. His work bridges theoretical mathematics with practical applications in computational science, data science, and machine learning. Le Gia's recent publications demonstrate a strong focus on numerical methods for PDEs on spheres, stochastic analysis, and machine learning applications. His work shows consistent progression from theoretical foundations to practical implementations, with increasing interdisciplinary applications in recent years. L. F. Guseman Prize in Mathematics, Texas A&M University (2003) As a dedicated academic mentor, Le Gia has supervised numerous PhD, Master's, and Honours students across computational mathematics and data science topics. He has secured significant research funding through ARC Discovery Projects including DP220101811 (2022-2024) and DP180100506 (2018-2020). Professionally, he serves as External Associate Editor for Frontiers in Applied Mathematics and Statistics , Secretary for ANZIAM's Computational Mathematics Group, and Co-chair of Mathematics of Computation and Optimisation (AustMS Special Interest Group).
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).
Leo Schwinn is a Lecturer at the Technical University of Munich (TUM) within the Department of Computer Science (I26), working in the Data Analytics and Machine Learning group supervised by Prof. Stephan Günnemann at the TUM School of Computation, Information and Technology. His research focuses on robust machine learning with particular emphasis on data-efficient learning and robustness vulnerabilities of Large Language Models (LLMs). Dr. Schwinn's research interests span multiple critical areas in contemporary machine learning including: Robustness against adversarial attacks in LLMs Embedding space vulnerabilities and defenses Model unlearning and privacy preservation Efficient training methodologies for large models Time-series forecasting with probabilistic frameworks Graph-based machine learning approaches His work bridges theoretical understanding with practical security implications of modern AI systems. Analysis of his recent publications (2023-2025) reveals a strong focus on LLM security, with multiple papers accepted at premier conferences including ICML, CVPR, ICLR, and NeurIPS. His research demonstrates consistent innovation in identifying novel attack vectors while developing practical defense mechanisms, particularly through embedding space manipulation techniques. The work shows increasing sophistication in handling both theoretical aspects of model robustness and practical deployment concerns. His notable scientific achievements include: Receiving the ATE dissertation price for his PhD work at FAU Securing an oral presentation at ICLR 2025 Organizing the ICLR BlogPost Track Becoming a member of ELLIS (European Laboratory for Learning and Intelligent Systems) Dr. Schwinn has served as review process chair for the 2024 Conference on Lifelong Learning Agents (CoLLAs) and actively collaborates with researchers at Mila Quebec AI Institute. His research group at TUM focuses on addressing fundamental challenges in machine learning robustness, particularly as they apply to real-world deployment scenarios where security and reliability are paramount. He maintains active GitHub repositories related to LLM security research, including circuit-breakers-eval and LLM_Embedding_Attack, demonstrating his commitment to open science and reproducible research in the field of AI security.
Geoff Pleiss is an Assistant Professor in the Department of Statistics at the University of British Columbia (UBC), affiliated with CAIDA's AIM-SI cluster. He is also a Canada CIFAR AI Chair and faculty member at the Vector Institute. His research bridges deep learning and probabilistic modeling, focusing on uncertainty quantification, Bayesian optimization, Gaussian processes, and ensemble methods. Pleiss earned his PhD in Computer Science from Cornell University (2020), followed by a postdoc at Columbia University. He holds multiple awards, including the AISTATS Top Reviewer and NeurIPS recognitions. His work emphasizes scalable algorithms and open-source contributions, such as the GPyTorch library. Pleiss advises students in Computer Science and Statistics, including Donney Fan (PhD), Tim G. Zhou (MSc), and others. He teaches advanced courses like STAT 547U (Deep Learning Theory) and STAT 520P (Bayesian Optimization). Grants include NSERC Discovery and New Frontiers in Research funding. Pleiss collaborates on interdisciplinary projects, such as astrophysical discovery via machine learning, and actively participates in academic service and outreach. Education: PhD in Computer Science, Cornell University (2020) MSc in Computer Science, Cornell University (2018) BSc in Engineering (Computing with Applied Mathematics), Olin College (2013) Key Research Themes: Uncertainty-aware decision-making with neural networks Scalable Gaussian processes and Bayesian optimization Ensemble methods and their theoretical limitations Recent Grants: NSERC Discovery Grant (2024) New Frontiers in Research Fund (2025, co-PI) His publications span foundational theory to applied machine learning, with over 14,500 citations. He actively mentors students through research internships and advises on open-source software development. Pleiss frequently presents at top conferences and collaborates with industry partners like Microsoft and ASAPP.
Pekka Marttinen is a tenured Associate Professor of Machine Learning at Aalto University, Department of Computer Science, and leads the Machine Learning for Health (Aalto-ML4H) group within the Helsinki Institute for Information Technology HIIT. Education: M.Sc. in Applied Mathematics, University of Helsinki (2004) Ph.D. in Statistics, University of Helsinki (2008) Title of Docent in Information and Computer Science, Aalto University (2015) Research Focus: His methodological work spans large language models, reinforcement learning, deep learning, probabilistic machine learning, and causal inference . These techniques are applied to critical domains of healthcare, bioinformatics, statistical genetics, epidemiology, and personalized medicine . The group develops novel algorithms, theoretical guarantees, and open-source software that enable data-driven discovery and decision-making in medicine and biology. Publication Trends: Across 2022-2024 the lab has concentrated on (i) rigorous causal reasoning over temporal clinical data, (ii) principled uncertainty quantification in LLMs, (iii) representation learning for neural network comparison, and (iv) translational projects that turn raw EHRs into actionable clinical insights. Earlier work integrated high-dimensional genomics with metabolomics and mapped evolutionary forces in bacterial pathogens. Scientific Awards & Recognition: While no specific awards are listed, his sustained publication record in top-tier venues (NeurIPS, ICML, AISTATS, Nature Genetics, PLOS CB) and his role as responsible professor of the Machine-Learning, Data-Science and AI major signify significant peer recognition. Advising & Grants: Prof. Marttinen currently mentors 8 PhD students as primary supervisor and an additional 7 PhD students as co-supervisor. He has already graduated 8 PhDs since 2014. The group is supported through competitive funding including the Finnish Center for Artificial Intelligence (FCAI) doctoral program. Labs & Teams: He directs the Machine Learning for Health (Aalto-ML4H) research group, comprising postdocs Hans Moen, Ti John, Alexander Nikitin, Negar Safinianaini, Linli Zhang, Zhiyuan Li, and the above-mentioned PhD cohort.
