Fast Bayesian inference is great… until you’re babysitting convergence.

Alex Andorra is joined by Martin Ingram to explore DADVI a more predictable, less noisy approach to variational inference that makes trade-offs explicit instead of mysterious

🎧 https://lnkd.in/gAX2iaHz

#bayesianinference

New Bayesian method boosts quantum dot charge detection—faster, smarter quantum tech is here. #QuantumComputing #BayesianInference #NanoTech

https://geekoo.news/bayesian-breakthrough-in-quantum-dot-charge-detection/

Developing Bayesian inference methods for complex scientific problems?

#EuroSciPy2025 is seeking original work on Hamiltonian Monte Carlo, variational inference, and statistical modeling in #Python.

Submit your innovations: https://pretalx.com/euroscipy-2025/cfp #CfP

#BayesianStatistics #ScientificPython #BayesianInference #PyMC #PyStan #EuroSciPy

Weekly Update at the Open Journal of Astrophysics – 08/03/2025

Time for the weekly Saturday morning update of papers published at the Open Journal of Astrophysics. Since the last update we have published four new papers, which brings the number in Volume 8 (2025) up to 25 and the total so far published by OJAp up to 260.

In chronological order of publication, the four papers published this week, with their overlays, are as follows. You can click on the images of the overlays to make them larger should you wish to do so.

The first paper to report is “Partition function approach to non-Gaussian likelihoods: information theory and state variables for Bayesian inference” by Rebecca Maria Kuntz, Heinrich von Campe, Tobias Röspel, Maximilian Philipp Herzog, and Björn Malte Schäfer, all from the University of Heidelberg (Germany). It was published on Wednesday March 5th 2025 in the folder Cosmology and NonGalactic Astrophysics and it discusses the relationship between information theory and thermodynamics with applications to Bayesian inference in the context of cosmological data sets.

You can read the officially accepted version of this paper on arXiv here.

The second paper of the week  is “The Cosmological Population of Gamma-Ray Bursts from the Disks of Active Galactic Nuclei” by Hoyoung D. Kang & Rosalba Perna (Stony Brook), Davide Lazzati (Oregon State), and Yi-Han Wang (U. Nevada), all based in the USA. It was published on Thursday 6th March 2025 in the folder High-Energy Astrophysical Phenomena. The authors use models for GRB electromagnetic emission to simulate the cosmological occurrence and observational detectability of both long and short GRBs within AGN disks

You can find the officially accepted version of this paper on arXiv here.

The next two papers were published on Friday 7th March 2025.

“The distribution of misalignment angles in multipolar planetary nebulae” by Ido Avitan and Noam Soker (Technion, Haifa, Israel) analyzes the statistics of measured misalignment angles in multipolar planetary nebulae implies a random three-dimensional angle distribution limited to <60 degrees. It is in the folder Solar and Stellar Astrophysics.

Here is the overlay:

The official published version can be found on the arXiv here.

The last paper to report this week is “The DESI-Lensing Mock Challenge: large-scale cosmological analysis of 3×2-pt statistics” by Chris Blake (Swinburne, Australia) and 43 others; this is a large international collaboration and I apologize for not being able to list all the authors here!

This one is in the folder marked Cosmology and NonGalactic Astrophysics; it presents an end-to-end simulation study designed to test the analysis pipeline for the Dark Energy Spectroscopic Instrument (DESI) Year 1 galaxy redshift dataset combined with weak gravitational lensing from other surveys.

The overlay is here:

You can find the “final” version on arXiv here.

That’s all for this week. It’s good to see such an interesting variety of topics. I’ll do another update next Saturday

#3x2ptAnalysis #ActiveGalacticNuclei #arXiv241113625v2 #arXiv241212548v2 #arXiv241217714v2 #arXiv250104549v2 #BayesianInference #Cosmology #CosmologyAndNonGalacticAstrophysics #DESI #DiamondOpenAccess #DiamondOpenAccessPublishing #entropy #GammaRayBursts #HighEnergyAstrophysicalPhenomena #InformationTheory #numericalSimulations #planetaryNebulae #SolarAndStellarAstrophysics #StatisticalMechanics #WeakLensing

I'm explaining Hamiltonian Monte Carlo in my grad-level stats class tomorrow, so I put together this animation illustrating HMC in one dimension. I find it very soothing.

#bayesian #BayesianInference #posterior #stats #r #rlang #statistics #MCMC

I'm teaching my first lecture at the new job today, about probabilistic logic programming, probabilistic inference, and (weighted) model counting.

Some of the required reading is a paper (https://eccc.weizmann.ac.il/eccc-reports/2003/TR03-003/index.html) that was written by a great mentor of mine, prof. dr. Fahiem Bacchus. He passed away just over 2 years ago, and I am honoured to keep his memory alive by teaching his ideas to a new generation of students. Hope to do him proud. 🌱

Please send good vibes? 🥺

#AcademicChatter #AcademicLife #AcademicMastodon #Teaching #Probability #ProbabilisticInference #Probabilities #Logic #LogicProgramming #PropositionalModelCounting #ProbabilisticLogicProgramming #ModelCounting #PropositionalLogic #WeightedModelCounting #DPLL #BayesianProbability #BayesNets #BasianStatistics #BayesianInference #BayesianNetworks #KnowledgeCompilation #DecisionDiagrams #BinaryDecisionDiagrams

Two New Publications at the Open Journal of Astrophysics

It’s Saturday morning again so here’s another report on activity at the  Open Journal of Astrophysics.  Since the last update we have published two more papers, taking  the count in Volume 7 (2024) up to 95 and the total published by OJAp up to 210.  We’ve still got a few in the pipeline waiting for the final versions to appear on arXiv so I expect we’ll reach the 100 mark for 2024 in the next couple of weeks.

The first paper of the most recent pair, published on October 22 2024,  and in the folder marked Astrophysics of Galaxies, is “Cloud Collision Signatures in the Central Molecular Zone”  by Rees A. Barnes and Felix D. Priestley (Cardiff University, UK) .  This paper presents an analysis of combined hydrodynamical, chemical and radiative transfer simulations of cloud collisions in the Galactic disk and Central Molecular Zone (CMZ).

Here is a screen grab of the overlay which includes the abstract:

You can click on the image of the overlay to make it larger should you wish to do so.  You can find the officially accepted version of this paper on the arXiv here.

The second paper has the title “Partition function approach to non-Gaussian likelihoods: macrocanonical partitions and replicating Markov-chains” and was published October 25th 2024. The authors are Maximilian Philipp Herzog, Heinrich von Campe, Rebecca Maria Kuntz, Lennart Röver and Björn Malte Schäfe (all of Heidelberg University, Germany). This paper, which is in  the folder marked Cosmology and NonGalactic Astrophysics, describes a method of macrocanonical sampling for Bayesian statistical inference, based on the macrocanonical partition function, with applications to cosmology.

Here is a screen grab of the overlay which includes the abstract:

You can click on the image of the overlay to make it larger should you wish to do so. You can find the officially accepted version of the paper on the arXiv here.

That concludes this week’s update. More  next week!

#arXiv231116218v3 #arXiv240721575v2 #AstrophysicsOfGalaxies #BayesianInference #CosmologyAndNonGalacticAstrophysics #likelihoods #MarkovChains #MolecularCouds #PartitionFunction #starFormation #thermodynamics

It’s Saturday morning again so here’s another report on activity at the  Open Journal of Astrophysics.  Since the last update we have published two more papers, taking  the count in Volume 7 (2024) up to 73 and the total published by OJAp up to 188.  We’ve still got a few in the pipeline waiting for the final versions to appear on arXiv so I expect we’ll reach the 200 mark fairly soon.

The first paper of the most recent pair, published on September 4th 2024,  and in the folder marked Astrophysics of Galaxies, is “Massive Black Hole Seeds”  by John Regan of the Department of Theoretical Physics at Maynooth University and Marta Volonteri (Sorbonne Université, Paris, France). This article presents a discussion of the pathways to the formation of massive black holes, including both light and heavy initial seeds.

Here is a screen grab of the overlay which includes the abstract:

You can click on the image of the overlay to make it larger should you wish to do so. Those of you who are paying attention will see that there is a bit of a glitch on the left hand side where software has thrown a line break in between the two author names. I have no idea what caused this so I raised a ticket with Scholastica and no doubt it will soon be fixed.  (Update: it is now fixed, 12th September 2024). You can find the officially accepted version of this paper on the arXiv here.

The second paper has the title “The future of cosmological likelihood-based inference: accelerated high-dimensional parameter estimation and model comparison” and was published on 5th September 2024. The authors are Davide Piras (Université de Genève), Alicja Polanska (MSSL) , Alessio Spurio Mancini (Royal Holloway, London), Matthew A. Price(UCL) & Jason D. McEwen (UCL); the latter four are all based in the UK. This paper, which is in  the folder marked Cosmology and NonGalactic Astrophysics, describes an accelerated approach to Bayesian inference in higher-dimensional settings, as required for cosmology, based on recent developments in machine learning and its underlying technology.

Here is a screen grab of the overlay which includes the abstract:

You can click on the image of the overlay to make it larger should you wish to do so. You can find the officially accepted version of the paper on the arXiv here.

That concludes this week’s update. More  next week!

https://telescoper.blog/2024/09/07/two-new-publications-at-the-open-journal-of-astrophysics-15/

#arXiv240512965v2 #AstrophysicsOfGalaxies #BayesianInference #BlackHoleSeeds #blackHoles #CosmologyAndNonGalacticAstrophysics #DiamondOpenAccess #likelihoodBasedInference #OpenJournalOfAstrophysics #TheOpenJournalOfAstrophysics

New on the blog: showcasing the immense hackability of #brms by extending a random intercept model with linear predictors on the standard deviation of the random intercept. Should you do it? Most likely not, but if you really really want, there is a way. Also the techniques shown are general and let you do a lot of other crazy stuff with brms. Happy for any feedback!
https://www.martinmodrak.cz/2024/02/17/brms-hacking-linear-predictors-for-random-effect-standard-deviations/

#bayesian #BayesianStatistics #BayesianInference #MixedModels

Agora com a "moda" da Inteligencia Artificial muita gente me vem perguntar sobre o assunto.

Respondo apenas o basico que sei e evito botar faladura sobre o assunto.

Foi uma area que nao acompanhei/acompanho desde 1990, altura em que tive as ultimas cadeiras nessa area (4 se bem me recordo), pelo que sou ignorante no tema, em particular as evolucoes que teve nos ultimos 30 anos.

#lisp #scheme #prolog #expertsystems #neuralnetworks #turingtest #bayesianinference

Who'd like to work out the "likelihood" that all this is simply coincidence? 😛

William #Blake died in 1827, the same year as Pierre-Simon #LaPlace, and is buried in the same cemetery as Thomas #Bayes. In 1788 Blake wrote this:

"...the ratio of all we have already known, is not the same that it shall be when we know more."

#BayesianInference
#InverseProbability
#WilliamBlake
#PierreSimonLaPlace
#ThomasBayes
#Statistics

After a long break, new #arxivfeed

"Generalized Bayesian Inference for Scientific Simulators via Amortized Cost Estimation"
https://arxiv.org/abs/2305.15208

#BayesianInference #MachineLearning #Modelling #SBI #ComputationalNeuroscience

Happy to share that the second paper of my PhD is now available as preprint and open for public discussion:
https://doi.org/10.5194/egusphere-2023-222

We developed a stochastic model of regional surface heat flow and Bayesian methods for its quantification. In particular, we aim to infer the strength of a specifically shaped signal given a sample of heat flow measurements.
#geophysics #heatflow #openscience #BayesianInference

Our new paper is now out in #JCIM !

BICePs v2.0: Software for Ensemble Reweighting Using Bayesian Inference of Conformational Populations

https://doi.org/10.1021/acs.jcim.2c01296

Congrats to Rob Raddi on this paper and for coding this more user-friendly and extensible Python package #BayesianInference #compchem #chemistry #biophysics

"Dear all,

We are thrilled to announce the inaugural #ComputationalPsychiatry Conference to take place at Trinity College Dublin on July 6-8th, 2023 (#cpconf2023)

https://www.cpconf.org/

One of the key aims of #ComputationalNeuroscience is to construct theoretical accounts of normal mental function that link characterizations of #neurobiology, #psychology and aspects of the environment. In Computational Psychiatry (CP), these theories, realized in models at various scales, are used to elucidate dysfunction.

The 2023 Computational Psychiatry Conference (7th and 8th July) will contain six sessions, each with a keynote talk from senior faculty and also contributed talks and panel discussions.

The session themes will include Diagnostics, Reinforcement Learning models, Individual-level prediction, Development, Animal models and Treatments. There will also be poster sessions on both days.

The tutorial session (afternoon of 6th July) will contain three introductory talks on #psychiatry for non-clinicians, #BehaviouralModelling using #BayesianInference and #ReinforcementLearning, and #MachineLearning.

Abstract submissions will be closed on March 15th, 2023. We will be able to support 10 participants with a travel award based on a competitive review of their abstract submissions. Top submissions will also be invited as talks.

We look forward to seeing everyone in Dublin this summer!"

🚀 #AWS Fortuna is skyrocketing! 🚀 Just a few days, and so many GitHub stars and forks! ⭐️

Fortuna supports #ConformalPrediction, #BayesianInference and other methods for #UncertaintyQuantification in #DeepLearning.

Try it out and let us know!
https://github.com/awslabs/fortuna

In collaboration with @cedapprox, @andrewgwils and team.

#uncertainty #neuralnetworks #bayesian #conformal #calibration #jax #flax #python #opensource #library #machinelearning #ai

"Our results show that a Bayesian machine can be implemented in a system with distributed #memristors, performing computation
locally, and with min. energy movement, allowing the computation of #BayesianInference with an energy efficiency more than three orders of magnitude higher than a standard microcontroller unit. Due to its reliance on non-volatile memory, and its sole use of read ops, once [...] programmed, the system may be powered down anytime while regaining functionality instantly. "

"Your use / of Bayesian inference // in the history of slavery": Alfred H. Conrad, "The Economics of the Antebellum South", and Adrienne Rich's "Winterface". #111Words #AdrienneRich #Poetry #AlfredHConrad #EconomicHistory #BayesianInference #HistoryOfSlavery https://andrewjshields.blogspot.com/2022/12/your-use-of-bayesian-inference-in.html

Today, we open sourced Fortuna (https://github.com/awslabs/fortuna) a library for uncertainty quantification.
Deep neural networks are often overconfident and do not know what they don’t know. Quantifying the uncertainty in the predictions they make will help deploy deep learning more responsibly and more safely.
#responsibleAI #ConformalPrediction #BayesianInference #UncertaintyQuantification #deeplearning #opensource

Don't forget to submit to the #PyMCon web series

Submissions are due November 30th!

Details here: https://pymcon.com/cfp

We'd love to receive your submission. Feel free to reach out with additional questions!

First-time speakers are especially encouraged to apply!

#bayes #BayesianInference #pymc