**Fully funded postdoc and doctoral student positions in various topics including Bayesian modeling, probabilistic programming and workflows** with me and other professors in Aalto University and University of Helsinki, funded by Finnish Center for Artificial Intelligence @FCAI

See more topics, how to apply, and job details like salary at https://fcai.fi/we-are-hiring

#Bayes #Bayesian #ML #AI #Job #PostDoc #PhD

Hi, I'm using #JAGS (just another Gibbs sampler) and I need to setup a conditional with four conditions.
I don't think JAGS has a powerful ifelse statement to control flow (but I may be wrong). Some people pointed me to the step function but I don't get it 😰

Any idea to write a conditional structure for 4 conditions?? It should be trivial to do!
#bayes #bayesian #statistics

Jozsef Arato has translated Chapters 1- 5 of (my + @ShravanVasishth + Daniel Schad) Intro to Bayes for Cog Sci (https://vasishth.github.io/bayescogsci/)

You can find it here:
https://github.com/jozsarato/bayescogdat

#cogsci #Bayesian #pymc #python #bayes @cognition @cogsci

- There are several language feature depracations that turn to errors. Probably the most widely used is the old array syntax, e.g. `int arr[5];` gives an error. Use `array[5] int arr;` instead. For this version, all the deprecated features in your Stan code can be updated using the autoformatter. with `--canonicalize=deprecations`.

Please help testing the release candidate with your own models!

See more in https://discourse.mc-stan.org/t/cmdstan-stan-2-33-release-candidate/32531

#Bayes #Bayesian #Stan

2/2

CmdStan 2.33 release candidate is now available!

Two major new things are
- a new language type tuples, which allows, e.g., `eigendecompose()` to return both vectors and values with one function call

- a new variational inference algorithm Pathfinder (soon available also in interfaces such as CmdStanR)

#Bayes #Bayesian #Stan

1/2

#bayes #bayesian + #onepunchman #opm =

Anyone with a #Bayes background interested in reviewing a new package for the Journal Of Open Source Software? 😊

Ping me with your Github handle if so.

@dougfort #bayes

...It was Fisher who then discredited inverse probability, and because of his influence it remained discredited until the late 20th century, when Fisher's methods were finally recognized to be flawed instead (https://doi.org/10.1080/00031305.2016.1154108 and https://doi.org/10.1080/00031305.2019.1583913 their flaws, though, were known since the 1930s).

For the history from Laplace to Fisher see for instance Dale (https://archive.org/details/historyofinverse0000dale).

@dougfort #bayes

It's sad how important figures in the history of science get forgotten (often together with their insights).

The original piece
https://www.lindahall.org/about/news/scientist-of-the-day/john-canton
has some gross historical inaccuracies. After Laplace rediscovered and successfully used inverse probability (https://www.jstor.org/stable/41133662), it kept on being used until the beginning of the 1900s, not forgotten as claimed in the piece...

New version of cmdstanr exposes Stan functions and allows you to directly unconstrain parameters (super useful for seeing parameters as the sampler sees them)! A huge win!

https://mc-stan.org/cmdstanr/news/index.html#cmdstanr-060

#Stan #cmdstanr #mcmc_stan #Bayes #rstats

Since this is a #bayes instance, a little detail here:

I used to get shit for applying probabilistic models. "Why did they do X instead of Y please tell me!" but at least that seems to be on the decline.

This is what I got from #midjourney for the #Bayes Theorem.
No clue why the model came up with that but it's beautiful
#Math #Probability

1763: The work of Thomas Bayes introduced the need to use the base rate in determining a probability. #Poetry #Science #History #Probability #BayesTheorem #Bayes (https://sharpgiving.com/thebookofscience/items/p1763.html)

In this week's #ISE2023 lecture, we were discussing Naive Bayes Classification as a simple yet powerful algorithm used for solving classification problems.

#nlp #classification #Bayes #lecture #Aiart #stablediffusionart @fizise @KIT_Karlsruhe

Nice blog post on informative priors for correlation matrices: A joint prior tcombining LKJ (for positive semi-definiteness) and a normal prior (to inform the magnitude of individual correlations).

http://srmart.in/informative-priors-for-correlation-matrices-an-easy-approach/

@smartin2018, did you ever finish the short paper on this, which Sean mentions in the comments of the blog post?

#bayes #bayesian

Un mio articolo pubblicato qualche mese fa su Prisma è ora disponibile on line: "La #probabilità condizionata del reverendo #Bayes"

https://www.prismamagazine.it/2023/04/28/la-probabilita-condizionata-del-reverendo/

Language to talk about #uncertainty adopted by British defense. I wonder whether science could adopt some sort of similar language.

#datascience
#bayes

https://www.gov.uk/government/news/defence-intelligence-communicating-probability88

Approximating Naive Bayes on Unlabelled Categorical Data

https://openreview.net/forum?id=KpElM2S9pw

#classification #classifier #bayes

RT @NickTolley4
Our newest preprint is up! 🧠 🚀
https://www.biorxiv.org/content/10.1101/2023.04.17.537118v1

Here we show how to:
📌apply simulation based inference to large-scale neural network models
📌uncover the circuit level mechanisms of common MEG/EEG biomarkers with our lab's neural modeling software HNN.

#neuro #bayes

I'm looking for a good #OpenSource GUI tool to model Bayesian belief networks that is also reasonably accessible to beginners. Should run on Windows.

It's for a friend who isn't much into coding and would like to run some litigation decision analysis.

Any favorite tools/suggestions?

#Bayes #BeliefNetwork #Network #GUI #Litigation #RStats

@legaltech @law
@rstats