#𝗕𝗮𝘆𝗲𝘀𝗶𝗮𝗻
LightNews LightNews
tmlr-pub.bsky.social
Beyond Grids: Multi-objective Bayesian Optimization With Adaptive Discretization

Andi Nika, Sepehr Elahi, Cagin Ararat, Cem Tekin

Action editor: Branislav Kveton

https://openreview.net/forum?id=Wq150HaRVE

#optimizing #optimization #discretization
November 12, 2025 at 9:19 PM
metaarxivbot.bsky.social
Bias Corrected Vaccine Effectiveness in Test Negative Designs Using Serology Informed Bayesian Priors: A Reproducible FAIR/FHIR Native Pipeline: https://osf.io/ensxf
November 12, 2025 at 8:45 PM
lukemeister.bsky.social
Pleiades: A lot more than seven sisters!

arxiv.org/abs/2511.07533
Lost Sisters Found: TESS and Gaia Reveal a Dissolving Pleiades Complex Andrew W. Boyle, Luke G. Bouma, Andrew W. Mann Most star clusters dissolve into the Galaxy over tens to hundreds of millions of years after they form. While recent Gaia studies have honed our view of cluster dispersal, the exact chronology of which star formation events begat which star cluster remnants remains unclear. This problem is acute after 100 Myr, when cluster remnants have spread over hundreds of parsecs and most age estimates for main sequence stars are too imprecise to link the stars to their birth events. Here we develop a Bayesian framework that combines TESS stellar rotation rates with Gaia kinematics to identify diffuse remnants of open clusters. We apply our method to the Pleiades, which previous studies have noted shows kinematic similarities to other nearby young stellar groups. We find that the Pleiades constitutes the bound core of a much larger, coeval structure that contains multiple known clusters distributed over 600 pc. We refer to this structure as the Greater Pleiades Complex. On the basis of uniform ages, coherent space velocities, detailed elemental abundances, and traceback histories, we conclude that most stars in this complex originated from the same giant molecular cloud. This work establishes a scalable approach for tracing the genealogies of nearby clusters and further cements the Pleiades as a cornerstone of stellar astrophysics. We aim to apply this methodology to other associations as part of the upcoming TESS All-Sky Rotation Survey. Comments: Published in The Astrophysical Journal. Figure 15 is my favorite! Cite as: arXiv:2511.07533 [astro-ph.SR] (or arXiv:2511.07533v1 [astro-ph.SR] for this version) https://doi.org/10.48550/arXiv.2511.07533 Journal reference: ApJ, 994, 24 (2025) Related DOI: https://doi.org/10.3847/1538-4357/ae0724 Figure 15. The full extent of the Greater Pleiades Complex as it would appear on the night sky if all stars from our membership list were visible to the naked eye. The seven brightest stars in the Pleiades are shown in green, while the stars from our membership list are shown in white. The Big Dipper, Orion, and Taurus are overlaid in blue. Of the 3,091 stars in our membership catalog, 1,603 are above the horizon at the selected observing conditions. The plot is oriented to match what an observer looking to the South would see.
November 12, 2025 at 8:34 PM
arxiv-nucl-th.bsky.social
New paper: "Bayesian Analysis of the Neutron Star Equation of State and Model Comparison: Insights from PSR J0437+4715, PSR J0614+3329, and Other Multi-Physics Data", link: https://arxiv.org/abs/2509.07109
November 12, 2025 at 8:15 PM
anem0ne.net
i mean maybe it's just that virtually everything in here confirms my bayesian priors, but i don't think he's wrong at all

www.waleed-shahid.com/p/the-hetero...
The track record of “moderation” is less a theory than a list of decisions we now regret. The last rounds of triangulation weren’t costless experiments; they produced real harms. The list includes votes for the Iraq War; curbs on abortion access; opposition to marriage equality and DREAMers; more oil and gas drilling; a moral panic over an Islamic center in Lower Manhattan; the slow strangling of the public option. And even the “win by sounding different” theory runs into two hard constraints. First, if heterodoxy is more than tone, it has to win the party—carry a presidential primary and redefine “Democratic” from the top down. That is a tall order. The primary electorate is disproportionately liberal and heavily Black, deeply committed to climate, union, civil, and reproductive rights. A program that trims those commitments for “electability” will be rejected by the very voters asked to bear the cost. You don’t get a new brand without first winning the contest that creates the brand. Second, the idea that progressives in blue metros should simply quiet down is pure fantasy. Those cities set tomorrow’s middle because politics and policy get tested there—paid leave, childcare, rent control, $15 minimum wage, marriage equality, taking on ICE, decarbonization standards, small-donor systems, and decarceration. As Zohran Mamdani likes to say, it’s impossible until it isn’t. Heterodox strategists rarely acknowledge—let alone explain—that blue cities as laboratories of progressivism are part of the party’s heterodoxy too. Heterodoxy endures because it sells a comforting illusion: that a little Keir Starmer-style “hippy-punching” can substitute for rebuilding power. It offers a fantasy of politics where messaging tweaks can reverse decades of nationalization, where posture outweighs geography, and where a party that’s lost whole regions can win them back with the right tone. This is not just theoretical. In the U.K., Starmer’s sharp rebukes of progressives and insistence on a tougher immigration frame haven’t stemmed the rise of far-right Nigel Farage’s Reform UK — the populist wave he leads is now outpacing Labour in the polls and local elections.
November 12, 2025 at 8:04 PM
julienmayor.bsky.social
Our Bayesian analyses provided decisive evidence against the didactic effect, suggesting that parents do not adapt their speech based on their child individual word’s attributed knowledge. Twist: vowel dispersion increased for known words with age!

This may suggest reinforcement, not instruction 🤔
November 12, 2025 at 7:51 PM
paperposterbot.bsky.social
link 📈🤖
Bayesian inference for cluster-randomized trials with multivariate outcomes subject to both truncation by death and missingness (Tong, Li, Velazquez et al) Cluster-randomized trials (CRTs) on fragile populations frequently encounter complex attrition problems where the reasons for missing
November 12, 2025 at 7:05 PM
sdellicour.bsky.social
Bayesian phylogeographic inference is widely used in molecular epidemiological studies to reconstruct the dispersal history of pathogens. Discrete phylogeographic analysis treats geographic locations as discrete traits and infers lineage transition events among them, (3/9)
November 12, 2025 at 6:35 PM
soc-net-bot.bsky.social
Global Optimization on Graph-Structured Data via Gaussian Processes with Spectral Representations
https://arxiv.org/abs/2511.07734
Bayesian optimization (BO) is a powerful framework for optimizing expensive black-box objectives, yet extending it to graph-structured domains remains challenging...📈🤖
November 12, 2025 at 4:52 PM
d-m.bsky.social
In a Bayesian model, is "Sent from my iPad" a stronger or weaker indicator of the email's lameness than "This email may be privileged and/or confidential"?
November 12, 2025 at 4:51 PM
louisevans.bsky.social
it's fucking prion disease for bayesian jury systems
November 12, 2025 at 4:12 PM
minus9.bsky.social
Mike Godwin @sfmnemonic Drawing Bayesian inferences after extensive sampling, I've determined that it's 99-percent certain that anyone who uses "woke" as pejorative will turn out to be a fuckhead. Please don't blame me for pointing this out--it's just science.
November 12, 2025 at 4:08 PM
maosbot.bsky.social
My work is so influential that prospective PhD students are now being influenced by papers that an LLM thinks I wrote but actually did not
Your work on Gaussian processes for environmental modeling (“Active Learning of Environmental Models,” Osborne et al., 2012) and multi-objective optimization for sustainability trade-offs (“The Automation of Science,” King et al., 2009, with your contributions) directly informs my methodological approach. I propose: 1. Bayesian optimization to identify high-impact monitoring
November 12, 2025 at 2:26 PM
drewpendergrass.bsky.social
🛰️ Satellites and ground-based monitors can measure pollutants in the air but cannot infer their sources directly; pollutants may have blown downwind or reacted away. During my PhD, I developed CHEEREIO (cheere.io), free software which tracks pollution back to its source using Bayesian optimization.
Home
Free software for emission quantification
cheere.io
November 12, 2025 at 2:10 PM
savagerat.bsky.social
Yeah, I was just cracking a not very funny joke about Bayesian inference. Probably poor taste given people died, so apologies on two counts.
November 12, 2025 at 12:04 PM
benkawam.bsky.social
We then develop a Bayesian stats model that takes raw behavioural observations as input, and can recover the generative model’s parameters. We finally illustrate how to use our models to ask causal questions (eg effect of sex on behaviour)
Posterior estimates of average holding times for each combination of sexes (top) and corresponding contrasts (bottom) for a population of Assamese macaques.
November 12, 2025 at 11:56 AM
dznepapers.bsky.social
New paper alert! Bayesian analyses for research on Alzheimer's disease and related disorders-updating one's knowledge.: Teipel, Stefan; Lutz, Michael
Alzheimer's & dementia / Diagnosis, assessment & disease monitoring Vol. 17, no. 4, p. e70109
Bayesian analyses for research on Alzheimer's disease and related disorders-updating one's knowledge.
Teipel, Stefan; Lutz, Michael Alzheimer's & dementia / Diagnosis, assessment & disease monitoring Vol. 17, no. 4, p. e70109
pub.dzne.de
November 12, 2025 at 11:48 AM
teeeeeeeee.bsky.social
Bayesian inference and conformal prediction (prediction intervals) in nnetsauce v0.18.1

https://thierrymoudiki.github.io/blog/2024/03/18/python/conformal-and-bayesian-regression

#Techtonique #DataScience #Python #rstats #MachineLearning
Blog post image
November 12, 2025 at 11:31 AM
savagerat.bsky.social
Nobody had been able to calculate the chances of the Bayesian sinking given its location and mast height.
November 12, 2025 at 11:28 AM
gsimpson.bsky.social
🌟 plot.gam() also has a new theme: scheme=2

This draws plots with 68 & 95% intervals (by default) and has a ggplot2-like grey plot background
Example plot showing the partial effects of the estimated smooths in a GAM with terms y ~ x0 + s(x1) + s(x2) + x(x3). Three panels are shown, one per smooth. The first panel is for s(x1) and shows an increasing, slightly curved function that increases slightly more as x1 increases. The second panel shows the partial effects for s(x2), which shows much more variation, being initially negative at x2=0, rising to a peak around x2=0.35, then falling to ~-1 at x2=0.5, then remaining roughly flat followed by a gradual decline thereafter. The final panel shows the partial effect of s(x3), which is ~0 for all x3, with uncertainty band covering 0 too. This predictor is known to have no effect in the simulated data. The uncertainty in the partial effects is shown by two credible interval bands; a dark blue central band is a 68% Bayesian credible interval, while the lighter blue outer interval is a 95% Bayesian credible interval. The background of each panel is light grey with white grid lines, in a similar style to ggplot2's default theme.
November 12, 2025 at 11:28 AM
gsimpson.bsky.social
🌟 plot.gam() gains a deriv argument, which if TRUE plots derivatives of univariate smooths instead of the usual partial effect plots

Partial effect plots can be confusing. With deriv=TRUE you see the change in Y (η) for a small change in X, which is comparable with usual interpretations of model 𝛽
Example plot showing the derivative of the estimated smooths in a GAM with terms y ~ x0 + s(x1) + s(x2) + x(x3). Three panels are shown, one per smooth. The first panel is for s(x1) and shows a positive derivative that increases slightly as x1 increases. The second panel shows the derivative for s(x2), which shows much more variation, being initially strongly positive from x2=0, moving gradually to strongly negative by x2=0.35, then back to 0 at x2=0.5, then remaining roughly flat thereafter. This reflects the strongly peaked nature of the fitted smooth s(x2). The final panel shows the derivative of s(x3), which is ~0 for all x3, with uncertainty band covering 0 too. This predictor is known to have no effect in the simulated data. The uncertainty in the derivatives is shown by two credible interval bands; a dark blue central band is a 68% Bayesian credible interval, while the lighter blue outer interval is a 95% Bayesian credible interval. The background of each panel is light grey with white grid lines, in a similar style to ggplot2's default theme.
November 12, 2025 at 11:28 AM
gsimpson.bsky.social
A new release of the mgcv #RStats 📦 is out on CRAN and Simon Wood (U Edinburgh) has added some significant new features despite the small bump in version number:

🌟 scasm() for estimating GAMs with shape constrained smooths. Can be used with any family & smoothness selection is via the EFS method
Model is: b3 <- scasm( y ~ s(x0, bs = "bs", k= k) + s(x1, bs = "sc", xt = "m+", k = k) + s(x2, bs = "bs", k = k) + s(x3, bs = "bs", k = k), family=poisson, bs=200 ) The second smooth `s(x1) is a shape constrained smooth with a positive monotonicity constraint (xt = "m+"). The `bs = 200` arguments uses 200 boostrap samples, which generates bootstrap distributions for each coefficient in the model. These bootstrap samples respect the shape constraints, while the usual +/- 2 SE credible intervals may not. The uncertainty in the partial effects is shown by two credible interval bands; a dark blue central band is a 68% Bayesian credible interval, while the lighter blue outer interval is a 95% Bayesian credible interval. The background of each panel is light grey with white grid lines, in a similar style to ggplot2's default theme.
November 12, 2025 at 11:28 AM
lunardial.nexus
bro really tried to do a top down bayesian hierarchical model of reading but in all of the wrong ways
Goodman concluded that: Skill in reading involves not greater precision, but more accurate first guesses based on better sampling techniques, greater control over language structure, broadened experiences and increased conceptual development. As the child develops reading skill and speed, he uses increasingly fewer graphic cues.
November 12, 2025 at 11:23 AM
emmahaslett.bsky.social
Lest we forget, it's only been a year and a bit since the sinking of the Bayesian. There's been speculation that its massive (though not *as* massive) 72m mast was somewhat to blame for its very quick sinking, although it was also very much a case of it being in the wrong place at the wrong time.
November 12, 2025 at 11:11 AM
modeltransfer.bsky.social
We’re proud of our team member Edoardo Peruzzi, who presented “Harsanyi’s Legacy: Bayesian Games and the Transformation of Microeconomic Theory” last Friday at the HOPE Workshop at the Center for the History of Political Economy at Duke University!
@hopecenter.bsky.social

#FrontierResearch #ERCStG
November 12, 2025 at 10:40 AM