Proceedings Track
Full research papers of up to 9 content pages; archival, double-blind peer-reviewed, published in PMLR proceedings.
Machine Learning/CORE Unranked
ProbML 2026
Symposium on Probabilistic Machine Learning
ProbML 2026 is a symposium co-located with ICML 2026 in Seoul, South Korea, focused on advancing the theory and practice of probabilistic machine learning, particularly Bayesian methods. It emphasizes both foundational research in probabilistic inference and novel applications in healthcare and climate change, fostering collaboration between methodological and applied researchers.
Key deadlines
Verified September 9, 2026Full paper
March 21, 2026
AoE
Conference timeline
Submission and decisions
Full paperKey deadline
March 21, 2026 · AoE
Paper fit
Contribution paths
A strong submission should clearly identify its contribution and evaluate it appropriately.
Workshop Track
Extended abstracts of 3-5 content pages; non-archival, double-blind reviewed, presented as posters.
Fast Track
Papers recently accepted at major ML venues (e.g., ICML 2025, NeurIPS 2025, JMLR); non-archival, no anonymization required, presented as posters.
Research areas in scope
01
Probabilistic and/or Bayesian ML Methods
Probabilistic and approximate inference: variational inference, Monte Carlo methods, expectation propagation, Laplace approximations, normalizing flowsModel development: Bayesian neural networks, Gaussian processes, probabilistic graphical models, state-space models, hierarchical models, and structured probabilistic modelsUncertainty quantification: calibration, out-of-distribution detection, uncertainty decomposition, conformal predictionTheory: convergence guarantees, approximation quality, PAC-Bayesian theoryDecision-making under uncertainty: Bayesian optimization, active learning, experimental design, multi-armed bandits, and distributional/Bayesian reinforcement learning, planning and search under uncertaintyScalability and efficiency: amortized methods, federated learningConnections to modern ML: Bayesian deep learning, uncertainty in foundation models, and uncertainty-aware generative AI
02
Applications of Probabilistic and/or Bayesian Methods with a focus on Healthcare and Climate Change
Clinical and public health applications: diagnosis and prognosis under uncertainty, individualized treatment recommendations, clinical trial design and monitoring, health policy evaluation, infectious disease modeling, resource allocation in healthcare systemsClimate and environmental applications: climate model calibration and emulation, extreme event modeling, uncertainty quantification in climate and weather projections, climate-informed decision support, environmental risk assessment, integration of physical and data-driven modelsDecision-making under uncertainty: Bayesian decision analysis, optimal experimental design, adaptive policies, decision support tools for practitioners and policymakersModel critique, validation, and robustness in real systems: model misspecification, distributional shifts, reliability in deployment, principled communication of uncertainty to stakeholders
Policies worth checking twice
- Submissions must be anonymized for Proceedings and Workshop tracks; author names must not reveal identity.
- Proceedings and Workshop track submissions must not be under consideration elsewhere during review; dual submission to other conferences or journals is prohibited.
- Extended abstracts and preprints (e.g., arXiv) are not considered concurrent submissions.
- For the Fast Track, papers must have been accepted at major ML venues after AABI 2025 and must follow their original camera-ready formatting.
- Supplementary material may be included but reviewers are not required to read it.
- Preprints must not be explicitly identified as ProbML submissions during the review process.
- Proceedings track papers must be camera-ready by 19 June 2026.
- Fast Track submissions are not archived and do not appear on the ProbML website.
Official sources
Compiled from the official call for papers. The organizers’ pages remain authoritative.
Last verified September 9, 2026