sigmoid.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
A social space for people researching, working with, or just interested in AI!

Server stats:

596
active users

#BayesianInference

0 posts0 participants0 posts today
EuroSciPy<p>Developing Bayesian inference methods for complex scientific problems?</p><p><a href="https://fosstodon.org/tags/EuroSciPy2025" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>EuroSciPy2025</span></a> is seeking original work on Hamiltonian Monte Carlo, variational inference, and statistical modeling in <a href="https://fosstodon.org/tags/Python" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Python</span></a>.</p><p>Submit your innovations: <a href="https://pretalx.com/euroscipy-2025/cfp" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="">pretalx.com/euroscipy-2025/cfp</span><span class="invisible"></span></a> <a href="https://fosstodon.org/tags/CfP" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>CfP</span></a></p><p><a href="https://fosstodon.org/tags/BayesianStatistics" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>BayesianStatistics</span></a> <a href="https://fosstodon.org/tags/ScientificPython" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ScientificPython</span></a> <a href="https://fosstodon.org/tags/BayesianInference" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>BayesianInference</span></a> <a href="https://fosstodon.org/tags/PyMC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>PyMC</span></a> <a href="https://fosstodon.org/tags/PyStan" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>PyStan</span></a> <a href="https://fosstodon.org/tags/EuroSciPy" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>EuroSciPy</span></a></p>
Peter McMahan<p>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.</p><p><a href="https://mas.to/tags/bayesian" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>bayesian</span></a> <a href="https://mas.to/tags/BayesianInference" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>BayesianInference</span></a> <a href="https://mas.to/tags/posterior" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>posterior</span></a> <a href="https://mas.to/tags/stats" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>stats</span></a> <a href="https://mas.to/tags/r" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>r</span></a> <a href="https://mas.to/tags/rlang" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>rlang</span></a> <a href="https://mas.to/tags/statistics" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>statistics</span></a> <a href="https://mas.to/tags/MCMC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>MCMC</span></a></p>
Dr. Anna Latour<p>I'm teaching my first lecture at the new job today, about probabilistic logic programming, probabilistic inference, and (weighted) model counting.</p><p>Some of the required reading is a paper (<a href="https://eccc.weizmann.ac.il/eccc-reports/2003/TR03-003/index.html" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">eccc.weizmann.ac.il/eccc-repor</span><span class="invisible">ts/2003/TR03-003/index.html</span></a>) 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. 🌱 </p><p>Please send good vibes? 🥺 </p><p><a href="https://mathstodon.xyz/tags/AcademicChatter" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AcademicChatter</span></a> <a href="https://mathstodon.xyz/tags/AcademicLife" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AcademicLife</span></a> <a href="https://mathstodon.xyz/tags/AcademicMastodon" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AcademicMastodon</span></a> <a href="https://mathstodon.xyz/tags/Teaching" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Teaching</span></a> <a href="https://mathstodon.xyz/tags/Probability" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Probability</span></a> <a href="https://mathstodon.xyz/tags/ProbabilisticInference" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ProbabilisticInference</span></a> <a href="https://mathstodon.xyz/tags/Probabilities" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Probabilities</span></a> <a href="https://mathstodon.xyz/tags/Logic" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Logic</span></a> <a href="https://mathstodon.xyz/tags/LogicProgramming" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LogicProgramming</span></a> <a href="https://mathstodon.xyz/tags/PropositionalModelCounting" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>PropositionalModelCounting</span></a> <a href="https://mathstodon.xyz/tags/ProbabilisticLogicProgramming" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ProbabilisticLogicProgramming</span></a> <a href="https://mathstodon.xyz/tags/ModelCounting" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ModelCounting</span></a> <a href="https://mathstodon.xyz/tags/PropositionalLogic" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>PropositionalLogic</span></a> <a href="https://mathstodon.xyz/tags/WeightedModelCounting" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>WeightedModelCounting</span></a> <a href="https://mathstodon.xyz/tags/DPLL" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>DPLL</span></a> <a href="https://mathstodon.xyz/tags/BayesianProbability" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>BayesianProbability</span></a> <a href="https://mathstodon.xyz/tags/BayesNets" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>BayesNets</span></a> <a href="https://mathstodon.xyz/tags/BasianStatistics" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>BasianStatistics</span></a> <a href="https://mathstodon.xyz/tags/BayesianInference" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>BayesianInference</span></a> <a href="https://mathstodon.xyz/tags/BayesianNetworks" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>BayesianNetworks</span></a> <a href="https://mathstodon.xyz/tags/KnowledgeCompilation" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>KnowledgeCompilation</span></a> <a href="https://mathstodon.xyz/tags/DecisionDiagrams" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>DecisionDiagrams</span></a> <a href="https://mathstodon.xyz/tags/BinaryDecisionDiagrams" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>BinaryDecisionDiagrams</span></a></p>
Martin Modrák<p>New on the blog: showcasing the immense hackability of <a href="https://fediscience.org/tags/brms" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>brms</span></a> 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!<br><a href="https://www.martinmodrak.cz/2024/02/17/brms-hacking-linear-predictors-for-random-effect-standard-deviations/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">martinmodrak.cz/2024/02/17/brm</span><span class="invisible">s-hacking-linear-predictors-for-random-effect-standard-deviations/</span></a></p><p><a href="https://fediscience.org/tags/bayesian" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>bayesian</span></a> <a href="https://fediscience.org/tags/BayesianStatistics" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>BayesianStatistics</span></a> <a href="https://fediscience.org/tags/BayesianInference" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>BayesianInference</span></a> <a href="https://fediscience.org/tags/MixedModels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>MixedModels</span></a></p>
Malte Ziebarth<p>Happy to share that the second paper of my PhD is now available as preprint and open for public discussion:<br><a href="https://doi.org/10.5194/egusphere-2023-222" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">doi.org/10.5194/egusphere-2023</span><span class="invisible">-222</span></a></p><p>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.<br><a href="https://norden.social/tags/geophysics" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>geophysics</span></a> <a href="https://norden.social/tags/heatflow" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>heatflow</span></a> <a href="https://norden.social/tags/openscience" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>openscience</span></a> <a href="https://norden.social/tags/BayesianInference" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>BayesianInference</span></a></p>
Gianluca Detommaso<p>🚀 <a href="https://sigmoid.social/tags/AWS" class="mention hashtag" rel="tag">#<span>AWS</span></a> Fortuna is skyrocketing! 🚀 Just a few days, and so many GitHub stars and forks! ⭐️</p><p>Fortuna supports <a href="https://sigmoid.social/tags/ConformalPrediction" class="mention hashtag" rel="tag">#<span>ConformalPrediction</span></a>, <a href="https://sigmoid.social/tags/BayesianInference" class="mention hashtag" rel="tag">#<span>BayesianInference</span></a> and other methods for <a href="https://sigmoid.social/tags/UncertaintyQuantification" class="mention hashtag" rel="tag">#<span>UncertaintyQuantification</span></a> in <a href="https://sigmoid.social/tags/DeepLearning" class="mention hashtag" rel="tag">#<span>DeepLearning</span></a>. </p><p>Try it out and let us know! <br /><a href="https://github.com/awslabs/fortuna" target="_blank" rel="nofollow noopener" translate="no"><span class="invisible">https://</span><span class="">github.com/awslabs/fortuna</span><span class="invisible"></span></a></p><p>In collaboration with <span class="h-card" translate="no"><a href="https://sigmoid.social/@cedapprox" class="u-url mention">@<span>cedapprox</span></a></span>, <span class="h-card" translate="no"><a href="https://sigmoid.social/@andrewgwils" class="u-url mention">@<span>andrewgwils</span></a></span> and team. </p><p><a href="https://sigmoid.social/tags/uncertainty" class="mention hashtag" rel="tag">#<span>uncertainty</span></a> <a href="https://sigmoid.social/tags/neuralnetworks" class="mention hashtag" rel="tag">#<span>neuralnetworks</span></a> <a href="https://sigmoid.social/tags/bayesian" class="mention hashtag" rel="tag">#<span>bayesian</span></a> <a href="https://sigmoid.social/tags/conformal" class="mention hashtag" rel="tag">#<span>conformal</span></a> <a href="https://sigmoid.social/tags/calibration" class="mention hashtag" rel="tag">#<span>calibration</span></a> <a href="https://sigmoid.social/tags/jax" class="mention hashtag" rel="tag">#<span>jax</span></a> <a href="https://sigmoid.social/tags/flax" class="mention hashtag" rel="tag">#<span>flax</span></a> <a href="https://sigmoid.social/tags/python" class="mention hashtag" rel="tag">#<span>python</span></a> <a href="https://sigmoid.social/tags/opensource" class="mention hashtag" rel="tag">#<span>opensource</span></a> <a href="https://sigmoid.social/tags/library" class="mention hashtag" rel="tag">#<span>library</span></a> <a href="https://sigmoid.social/tags/machinelearning" class="mention hashtag" rel="tag">#<span>machinelearning</span></a> <a href="https://sigmoid.social/tags/ai" class="mention hashtag" rel="tag">#<span>ai</span></a></p>
Marcel Fröhlich<p>&quot;Our results show that a Bayesian machine can be implemented in a system with distributed <a href="https://sigmoid.social/tags/memristors" class="mention hashtag" rel="tag">#<span>memristors</span></a>, performing computation<br />locally, and with min. energy movement, allowing the computation of <a href="https://sigmoid.social/tags/BayesianInference" class="mention hashtag" rel="tag">#<span>BayesianInference</span></a> 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. &quot;</p>
Cédric Archambeau<p>Today, we open sourced Fortuna (<a href="https://github.com/awslabs/fortuna" target="_blank" rel="nofollow noopener" translate="no"><span class="invisible">https://</span><span class="">github.com/awslabs/fortuna</span><span class="invisible"></span></a>) a library for uncertainty quantification.<br />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.<br /><a href="https://sigmoid.social/tags/responsibleAI" class="mention hashtag" rel="tag">#<span>responsibleAI</span></a> <a href="https://sigmoid.social/tags/ConformalPrediction" class="mention hashtag" rel="tag">#<span>ConformalPrediction</span></a> <a href="https://sigmoid.social/tags/BayesianInference" class="mention hashtag" rel="tag">#<span>BayesianInference</span></a> <a href="https://sigmoid.social/tags/UncertaintyQuantification" class="mention hashtag" rel="tag">#<span>UncertaintyQuantification</span></a> <a href="https://sigmoid.social/tags/deeplearning" class="mention hashtag" rel="tag">#<span>deeplearning</span></a> <a href="https://sigmoid.social/tags/opensource" class="mention hashtag" rel="tag">#<span>opensource</span></a></p>
Solal Nathan<p>🤔 Bayesian Inference (on graphical models) is NP-hard.</p><p>But even worst! every epsilon-approximation is also NP-hard.</p><p>Which means that the worst case scenario is (almost certainly) exponential.</p><p>Good news is, there are some special cases where approximation or exact inference can be performed efficiently.</p><p>📘 Check out more in &quot;Probabilistic Graphical Models: Principles and Technique&quot; by Daphne Koller and Nir Friedman</p><p><a href="https://sigmoid.social/tags/Bayes" class="mention hashtag" rel="tag">#<span>Bayes</span></a> <a href="https://sigmoid.social/tags/bayesianism" class="mention hashtag" rel="tag">#<span>bayesianism</span></a> <a href="https://sigmoid.social/tags/MachineLearning" class="mention hashtag" rel="tag">#<span>MachineLearning</span></a> <a href="https://sigmoid.social/tags/AI" class="mention hashtag" rel="tag">#<span>AI</span></a> <a href="https://sigmoid.social/tags/ML" class="mention hashtag" rel="tag">#<span>ML</span></a> <a href="https://sigmoid.social/tags/BayesianInference" class="mention hashtag" rel="tag">#<span>BayesianInference</span></a> <a href="https://sigmoid.social/tags/Inference" class="mention hashtag" rel="tag">#<span>Inference</span></a></p>