Dennis Prangle

New preprint: SBI with foundation models!
Tired of training or tuning your inference network, or waiting for your simulations to finish? Our method NPE-PF can help: It provides training-free simulation-based inference, achieving competitive performance with orders of magnitude fewer simulations! ⚡️

Paper on normalizing flow transformations for heavy tails 🦕 arxiv.org/abs/2406.16971 My co-author Tennessee Hickling will be presenting the poster at ICML icml.cc/virtual/2025...

Join our Cloud HD Video Meeting Zoom is the leader in modern enterprise cloud communications.

The next OWABI Seminar will take place during BayesComp 2025 in Singapore, on Thursday 19 June at 8pm Singapore time (1pm in Tórshavn). Speakers are: Filippo Pagani and Maurizio Filippone. Link to join: monash.zoom.us/j/810...
Meeting ID: 810 5099 4376
Passcode: 137607

The next OWABI seminar is quickly approaching, with two talks live from BioInference 2025 (bioinference.github....) on 29th May,10am UK time. Speakers: Andrew Golightly, Henrik Häggström. Links to the mailing list, to receive the Teams link to the talks, can be found on warwick.ac.uk/fac/sc...

I come back to this every few years and always get more out of it!

Election leaflet "West of England Mayor Election: It's Green or the Reform Party" with a bar chart showing Green 24%, Reform 20%, Labour 20%

When your analysis uses a very strong prior

ISBA 2026: Call for Invited Sessions & Panel Discussions The Scientific Committee invites the submission of proposals for Invited Sessions & Panel Discussions for the 2026 ISBA World Meeting, to be held in Nagoya, Japan from 28 June to 3 July, 2026. Invited...

The Scientific Committee welcomes proposals for Invited Sessions for the 2026 #ISBA World Meeting. Please submit proposals using this form: forms.gle/5G3xJWJaaE6Q...

Proposals will be accepted until 11:59pm (anywhere on Earth) on 6 July 2025.

Also see the conf. site: isba2026.github.io

Scientists on Bluesky, here's one of the best feeds I've found here:

Paper Skygest - a feed of posts containing research papers shared by people you follow.

The next OWABI seminar www.warwick.ac.uk/owabi is quickly approaching, being scheduled on Thursday the 27th March at 11am UK time. Our next speaker is Meïli Baragatti (Université de Montpellier), who will talk about "Approximate Bayesian Computation with Deep Learning and Conformal Prediction".

I will be one of the supervisors on this. Please pass on to any UK students who may be interested.

JOB OPENING at University of Bristol:
-- Lecturer in Statistics or Machine Learning
-- (x 2 positions available)
-- application deadline: 31 March, 2025

Details at www.bristol.ac.uk/jobs/find/de...

Come join us!

Our next OWABI talk will be given by Meïli Baragatti (Assistant Professor Institut Agro Montpellier) on Approximate Bayesian Computation with Deep Learning and Conformal prediction on Thursday the 27th March 2025. More info will follow.

I think there's an OnlyPosts feed you can use that excludes reposts. I also like the QuietPosters feed to highlight stuff that would get drowned out otherwise.

Our next talk will be on Thursday the 27th February at 11am UK time. Our next speaker is Ayush Bharti (Aalto University), who will talk about "Cost-aware simulation-based inference". To receive the link, sign up here: listserv.csv.warwick...

The deadline for abstract submission of contributed papers and posters for BayesComp2025 has been extended to 28 February. Decisions by 14 March. Submit here! bayescomp2025.sg/abstract-sub...
The deadline for early bird registration has been extended to March 22. Hope to see you in Singapore!

Senior Lecturer or Reader in Computational Statistics (Teaching & Research) :Oxford Road

Come and work with us! "Senior Lecturer or Reader in Computational Statistics ... biomedical data analysis, computational statistics and machine learning, foundations of data science and AI, high-dimensional data analysis, and uncertainty quantification."

www.jobs.manchester.ac.uk/Job/JobDetai...

Accelerating statistical inference and experimental design with machine learning - Isaac Newton Institute This workshop focuses on leveraging modern machine learning to accelerate statistical inference, experimental design, and scientific discovery. It features...

The Isaac Newton Institute at Cambridge is hosting a workshop on Accelerating Statistical Inference and Experimental Design with Machine Learning June 23 - 27. It includes tutorials on #Bayesian inference, DoE, and diffusion models.

More details at the website:

www.newton.ac.uk/event/rclw03/

Hello bluesky 🦋 - overdue intro post. I'm a statistician at Bristol University doing approx Bayes methods like SBI and variational influence for things like population genetics 🧬 and stochastic kinetic models. Happy to chat about scaling these methods to harder models/bigger data/real applications.

Invitation to review a revised paper I assumed wasn't coming back just days after I recycled my printout and written notes on the previous version. Maybe I should never tidy my office.

Petition to form a Bayesian Social Sciences section at ISBA Petition to form a Bayesian Social Sciences section at ISBA The following members of the International Society for Bayesian Analysis (ISBA) wish to petition the ISBA board to be recognized as the IS...

Bayesians!
Are you a member of ISBA? Are you into applications in the Social Sciences?

We are starting a new ISBA section on Bayesian Social Sciences. This requires a petition, signed by at least 30 ISBA members. You can add your name here: docs.google.com/document/d/1...

The Statistical Accuracy of Neural Posterior and Likelihood Estimation Neural posterior estimation (NPE) and neural likelihood estimation (NLE) are machine learning approaches that provide accurate posterior, and likelihood, approximations in complex modeling scenarios, ...

Thrilled to contribute to this work led by David Frazier providing theory for NPE/NLE in simulation-based inference. These methods are known to match the accuracy of ABC and BSL with fewer simulations, this paper rigorously shows why this can be achieved.
arxiv.org/abs/2411.12068

Reposting now there's more people here!

Thanks Umberto!

I created a starter pack for simulation-based inference (aka. likelihood-free inference).

Let me know if you’d like me to add you.

go.bsky.app/GVnJRoK

We've already submitted that paper elsewhere, but it could go here if they reject it!

Exciting upcoming special issue of Extremes on heavy tails 🦕and AI 🤖 link.springer.com/collections/...

GitHub - bayesflow-org/bayesflow at dev A Python library for amortized Bayesian workflows using generative neural networks. - GitHub - bayesflow-org/bayesflow at dev

Our Python library BayesFlow implements methods for amortized Bayesian inference. You first train a neural network on simulated data. Then you obtain posterior inference on any real data almost instantly. Check out the dev branch for our new backend and user interface: github.com/bayesflow-or...

The first talk of Season 6 is next week! On 31st October, 11am GMT, Ullrich Köthe (University of Heidelberg) will talk about "Free-form flows for physics-informed generative modelling". The talk will be on MS teams. Sign up to the list to get the link: listserv.csv.warwick...

This is the new account for the One World Approximate Bayesian Inference (ABI) Seminar. Please follow if you are interested in any flavour of approximate Bayesian inference, including simulation-based inference. Website: warwick.ac.uk/fac/sc...

Simple solution: use threads when I need a no-politics feed