Discover and read the best of Twitter Threads about #Bayesian

Most recents (24)

1. The impact case study (with .@MartinNeil9) on our #Bayesian network applications that was chosen as part of the .@QMUL #REF2021 submission achieved the highest possible rating 4*. Normally it’s not possible to know the rating for individual submissions...
2. Reason we know is because the QM Computer Science impact results were ranked joint top in the country – with 100% 4* ratings (so all 6 CS submissions were rated 4*). Here's a public summary of our case study from QM Press Office (apologies for typos!) eecs.qmul.ac.uk/research/featu…
3. The full submission included testimonies about critical applications with international organizations that cannot be made public because of confidentiality. The #Bayesian network software referred to is .@AgenaRisk agenarisk.com
Read 4 tweets
I've been studying #Bayesian methods in #rstats since the beginning of this year.

The more I learn, the more excited I get about Bayesian.

Here's why... Image
One of the key R packages I've been experimenting with is BRMS (Bayesian Regression Models using Stan).

BRMS allows us to model a wide range of statistical models including:

- linear,
- count data,
- survival,
- multi-effects,
- non-linear (& more!)
The important point is that Bayesian modeling implements a special technique called Markov Chain Monte Carlo (MCMC).

MCMC is a game changer.
Read 6 tweets
My biggest mistakes were never my insights. They were in over-confidence.

An #rstats + #bayesian 🧵
2/n In business, I've made great regression models that have predicted how much sales we were going to make.

In fact, this helped me increase revenue from $3M to $15,000,000 per year at one of the companies I worked at.

BUT my models were NOT perfect.
3/n In fact, I'd argue that the BIGGEST flops were due to over-confidence.

Believing my model was better than it actually was.

Here's what hurt me...
Read 6 tweets
New release for @Uber forecasting library 🌈

Orbit is an open-source #Python library for Bayesian time series forecasting and inference applications developed by @UberEng 🧵 👇🏼

#timeseries #forecast #MachineLearning #PyTorch #Bayesian #Bayes
The library uses under the hood probabilistic programming languages with libraries such as #Python @mcmc_stan , Pyro, and #PyTorch to build the forecast estimators.

The new release, version 1.1 includes the following new features and changes: 👇🏼
New #forecasting model - Kernel Time-based #Regression (KTR). KTR model uses latent variables to define a smooth, time-varying representation of regression coefficients. Tutorials: 👇🏼
orbit-ml.readthedocs.io/en/latest/tuto…
orbit-ml.readthedocs.io/en/latest/tuto…
orbit-ml.readthedocs.io/en/latest/tuto…
orbit-ml.readthedocs.io/en/latest/tuto…
Read 8 tweets
Did you know that An Introduction to Statistical Learning (ISLR) book has an online course? 🎥 🌈❤️
@edXOnline is offering an online course by @Stanford University, following the book curriculum: 🧵 👇🏼

#rstats #Statistics #ML #datascience
The course instructors are two of the book authors - Prof. Trevor Hastie and Prof. @robtibshirani. While the book is based on #R some awesome people translate it to #python, #julialang, and other #Rstats flavors (see links on the comments below 👇🏼).
The course covers the following topics (aligned with the book curriculum):
Read 9 tweets
New book for Bayesian statistics with #Python! 📚📊🚀
The Bayesian Modeling and Computation in Python by
@aloctavodia, @canyon289, and @junpenglao provides an introduction to Bayesian statistics using core Python libraries for Bayesian 🧵 👇🏼

#bayesian #MachineLearning #stats ImageImageImage
The book covers the following four topics:
- Bayesian Inference concepts
- Bayesian regression methods for linear regressions, splines,
- Time series #forecasting
- Bayesian additive regression trees
- Approximate of Bayesian computation
The book's authors are the contributors of PyMC3, ArviZ, Bambi, #TensorFlow Probability, and other #Python libraries:
PyMC3 - docs.pymc.io/en/v3/
Tensorflow Probability - tensorflow.org/probability
ArviZ - arviz-devs.github.io/arviz/
Read 6 tweets
Volume 100 of @jstatsoft: Software for Bayesian Statistics

Guest editors: @micameletti & @precariobecario

20 contributions on a wide range of methods #rstats #python #bayesian #inla @mcmc_stan #nimble #brms #rjags

URL: jstatsoft.org/v100 Editorial for the special volume 100 in the Journal of Stati
Van Niekerk, Bakka, Rue, Schenk:

New Frontiers in Bayesian Modeling Using the INLA Package in R

#rstats #inla #bayesian

doi.org/10.18637/jss.v…
Michaud, De Valpine, Turek, Paciorek, Nguyen:

Sequential Monte Carlo Methods in the nimble and nimbleSMC R Packages

#rstats #bayesian #nimble

doi.org/10.18637/jss.v…
Read 21 tweets
1/ Excited to share how T cell therapies kill #leukemia!! multi-omics + new #computational #singlecell tools for longitudinal analysis 👉unexpected answer! cell.com/cell-reports/f…

*👏* @elhamazizi! 🙏 @dpeer Cathy Wu @MDAndersonNews @CPRITTexas @ColumbiaBME @sloan_kettering
2/ We studied donor lymphocyte infusion (DLI) - an #Immunotherapy for relapsed #leukemia after #BMT & the #og of #celltherapy. Previously, we showed DLI reversed T cell exhaustion - but didn't know why/how/which T cells were responsible...
ashpublications.org/blood/article/…
3/ To address these ?'s, we modeled intraleukemic T cell dynamics by integrating longitudinal, multimodal data from ~100K T cells (!) during response (R) or resistance (NR: nonresponder) to DLI.
Read 18 tweets
Nebraska Attorney General's October 14 legal opinion on prescribing Ivermectin cites our #Bayesian analysis (page 16 & footnotes 101 and 104). Rules that it can be prescribed with informed consent ago.nebraska.gov/sites/ago.nebr… Image
2/3: This is the American Journal of Therapetics letter that was cited: dx.doi.org/10.1097/MJT.00…
3/3: And this is the full report that was cited:
arxiv.org/abs/2109.13739
arxiv.org/ftp/arxiv/pape…
Read 3 tweets
3. But the conclusions of such studies are also confounded by failing to consider non-Covid deaths; this overestimate the safety of the vaccine if there were serious adverse reactions. In fact multiple confounding factors will overestimate vaccine effectiveness.
4. One factor is how/whether a person is classified as a Covid ‘case’, Covid ‘hospitalization’ & Covid ‘death’. These can differ between vacc & unvaccinated. The unvaccinated who die ‘with’ as opposed to ‘from’ Covid are more likely to be classified as Covid deaths.
5. Another critical factor is how/whether a person is classified as ‘vaccinated’. Any person testing positive for Covid or dying of any cause within 14 days of their second dose is now classified by the CDC as ‘unvaccinated’
Read 14 tweets
(1/6)
I'm very happy that we (finally) got an acceptance for our submitted paper in the Journal of Open Source Education, JOSE🙂.

"An open source crash course on parameter estimation of computational models using a Bayesian optimization approach"
doi.org/10.21105/jose.…
(2/6)
Parameter estimation is a crucial aspect of model development in science and engineering. In the proposed educational module, we have a look at the #Bayesian optimization processes in general and model calibration (parameter estimation) in particular.
(3/6)
For demonstration purposes, we implement a model parameter estimation process for a fitting problem step by step in Python such that the readers can adapt it to their own models and use-cases.

Codes are available in this repo: github.com/mbarzegary/edu…
Read 7 tweets
Are you a part of this SME group, @davidcnorrismd?
Trial design and biostats folks, your expertise is needed. DM me for details on how to join this SME group. #biostats #trialdesign #bayesian #ChildhoodCancer
Read 5 tweets
If the connotation of risk is an intertwined concept and is difficult to quantify, how does a Risk Officer look at it?
Is there any way other than using copula models to determine systemic risk with long tails or a black swan event?
@CQFInstitute @GARP_Risk @SOActuaries
I guess we are worried about Market and Credit Risks or other interrelated financial risks which can create conjoint loss given events.
Any #Gaussian distribution model will enable you to model and predict potential Operational, Liquidity and Balance sheet AL - (Asset - liability) Mismatch, Market and Credit drove losses under normal market conditions.
Read 32 tweets
*NEW THREAD* This week we’ll be featuring the top 5 most impactful articles published in @RSeriesa in 2020 as captured by @altmetric

We’ll feature one paper per day in reverse order tarting by our 5th place

🧵👇🏽
1/n Image
Top 5th most impactful paper: Ellison et al: doi.org/10.1111/rssa.1…

*CONTEXT* Fertility projections are a key determinant of population forecasts, which are widely used by government policy makers and planners
2/n Image
*AIM* Propose an intuitive and transparent hierarchical #Bayesian model to forecast cohort fertility

*METHOD* Use of hamiltonian Monte Carlo methods and a data set from the human fertility database
3/n Image
Read 4 tweets
THREAD* We continue featuring the top 10 most cited articles published in @RSeriesa in 2020

Day 3: today we featured articles ranked 4th & 5th on the topics of #shipping 🚢 & #slavery

🧵starting👇🏽
1/n Image
The first article is by Silverman: doi.org/10.1111/rssa.1…

*CONTEXT* Quantifying hidden populations such as victims of modern slavery is challenging; yet, needed to develop strategies leading to legislation, inc. the Modern Slavery Act 2015
2/n Image
*AIM* Investigate the stability and robustness of various multiple‐systems‐estimate methods to measure modern slavery using Kosovo 🇽🇰as case study

*METHOD* Use of multiple‐systems‐estimate methods to develop a new Markov chain Monte Carlo #Bayesian approach
3/n
Read 8 tweets
My book recommendations for #Bayesian analysis:

• Kevin Murphy: probml.github.io/pml-book/book1…

• David MacKay: inference.org.uk/mackay/itila/

• David Barber: web4.cs.ucl.ac.uk/staff/D.Barber…

• Chris Bishop: microsoft.com/en-us/research…

All free online. Of course buy them if you can afford it 🙂
cc @fjsosah 👆
Others I hear good things about but haven't dug into as much (sorry):

@Cmrn_DP : ptgmedia.pearsoncmg.com/images/9780133…

@rlmcelreath xcelab.net/rm/statistical…

@willkurt : nostarch.com/learnbayes
Read 3 tweets
Thrilled to announce our new paper by Victoria Junquera and Adrienne Grêt-Regamey, "Assessing livelihood vulnerability using a Bayesian network: a case study in northern #Laos”, is now out in @ecologyandsociety1 ecologyandsociety.org/vol25/iss4/art… #openaccess #WomenInSTEM. Short thread👇
This work analyzes the effect of #cashcrop production on livelihood #vulnerability, which we define in terms of the probability distribution of household income.
We use a #Bayesian network to estimate the probability distribution of household income, conditional on biophysical (e.g., yield), household (e.g., agricultural land), and commodity price variables.
Read 10 tweets
Today's talk is with neuroscientist #KarlFriston of @ucl, who will offer a heuristic proof suggesting that life is an inevitable emergent property of any weakly mixing random dynamical system that possesses a #Markov blanket.

Tune in here at 12:15 PM MT:
santafe.edu/events/me-and-…
...and because it's 2020 and nothing is ever simple, we are having technical issues with the stream. The talk is recorded and we will upload it across all of our platforms ASAP. Our apologies!
Spatial boundaries are statistical boundaries: #KarlFriston on #MarkovBlankets, the reciprocal interfacing between internal and external states.

Follow this thread for more highlights from the talk and stay tuned for the video link...
Read 11 tweets
Chapter 4: Head rotation sensation is a splendid example of dynamic #Bayesian multisensory fusion since it involves several sensors with different dynamics. These sensor can be put in conflict or switched on/ off experimentally. Follow the tour! #vestibular
2/ We have (at least) 3 rotations sensors with different dynamics: the inner ear's canals detect acceleration; vision velocity, and graviceptors position (when rotating in vertical planes). The brain also relies on a zero velocity prior. Looks like a job for a #Kalmanfilter!
3/ I will explain (&simulate) motion perception during constant velocity rotations (starting from 0 velocity at t=0). Each sensor can report the motion, or 0, or be off altogether. There's experiments in the literature covering nearly all combinations! This will be a long thread!
Read 29 tweets
(1/n) I’ve been following discussions of #Bayesian sequential analyses, type I error, alpha spending etc. At the risk of offending everyone (please be kind!), I see reasons Bayesian sponsors and regulators can still find value in type I error rates and so forth.
(2/n) I’m focused on the design phase. After the trial, the data is the data. Lots of good stuff has been written on the invariance of Bayesian rules to stopping decisions. But in prospectively evaluating a trial design, even for a Bayesian there is a cost to interims, etc.
(3/n) Example…ultra simple trial. Normally distributed data. Known sigma=2. Mean is either 0 (null) or 1 (good). Simple decision space at end of trial…approve or not. A Bayesian utility would place values over the 4 combinations of truth/decision.
Read 14 tweets
Chapter 1: Why do we feel #dizzy when turning? This is because of how out inner ear’s rotation sensors (#vestibular semi-circular canals) work, from a mechanical point of view. Watch these movies and the next for explanations.
2/ The inner ear's #vestibular semi-circular canals are liquid-filled tubes. When the head rotates, the liquid stays in place and flows in the canal. This activates hair cells (in a structure called cupula) that sense the rotation.
3/ However, when turning too much, the liquid starts to rotate with the canal and the rotation signal fades out. Furthermore, when the rotation stops, the liquid keeps flowing and creates a rotation after-effect.
Read 10 tweets
~ New Post ~

During this quarantine time, I binge-watched @Stanford #CS330 lectures taught by the brilliant @chelseabfinn. This blog post is a summary of the key takeaways on #Bayesian Meta-Learning that I’ve learned. #AtHomeWithAI

medium.com/cracking-the-d…

(1/7) 👇
Bayesian meta-learning generates hypotheses about the underlying function, samples from the data distribution, and reasons about model uncertainty. It is suitable for problems in safety-critical domains, exploration strategies for meta-RL, and active learning.

(2/7) 👇
Read 7 tweets
Ever wonder what a bird sees? If so, read on!

Purple has always been my favorite color. But purple's not just pretty--it's a special type of color. If you look at a rainbow, you won't see purple. That's because purple is what we color scientists call a 'nonspectral' color. 1/n
Instead of being formed by a single light or mixture of similar lights, purple is formed by simultaneously stimulating our non-adjacent red-sensitive and blue-sensitive cones in our eyes. 2/n
Since purple is my favorite color, & birds are my favorite animals, I wondered, "Can birds see purple?" Today, our paper led by Dr. Cassie Stoddard with @dylanhmorris @B_G_Hogan @DavidInouye1 Ed Soucy is in @PNASNews that shows that yes, birds can! pnas.org/content/early/… 3/n
Read 16 tweets
Happy to finally see these results out! Project started 4 yrs ago from discussions with @noellebeckman at @sesync, study made possible with data from @stri_panama and resources from @UQAT & @ComputeCanada.
I would like to mention how technical challenges could be solved with the tools made available by the @mcmc_stan team, and @betanalpha's case studies illustrating best practices for Bayesian analysis: betanalpha.github.io/assets/case_st… . That story is mostly in Appendix 2 of our paper.
The main challenge in estimating #seeddispersal kernels is that they are often very leptokurtic, i.e. many seeds falling either very close or quite far from the parent plant. Seed sampling data has both an upper bound in terms of observed distances from parents...
Read 7 tweets

Related hashtags

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3.00/month or $30.00/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!