mastodon.world is one of the many independent Mastodon servers you can use to participate in the fediverse.
Generic Mastodon server for anyone to use.

Server stats:

8.1K
active users

#modelselection

2 posts2 participants0 posts today
Dr Mircea Zloteanu ☀️ 🌊🌴<p><a href="https://mastodon.social/tags/statstab" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>statstab</span></a> #393 Statistically Efficient Ways to Quantify Added Predictive Value of New Measurements [actual post]</p><p>Thoughts: #392 has the comments, but this is where the magic happens.</p><p><a href="https://mastodon.social/tags/modelselection" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>modelselection</span></a> <a href="https://mastodon.social/tags/modelcomparison" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>modelcomparison</span></a> <a href="https://mastodon.social/tags/variance" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>variance</span></a> <a href="https://mastodon.social/tags/effectsize" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>effectsize</span></a> <a href="https://mastodon.social/tags/tutorial" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>tutorial</span></a></p><p><a href="https://www.fharrell.com/post/addvalue/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="">fharrell.com/post/addvalue/</span><span class="invisible"></span></a></p>
Dr Mircea Zloteanu ☀️ 🌊🌴<p><a href="https://mastodon.social/tags/statstab" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>statstab</span></a> #392 Statistically Efficient Ways to Quantify Added Predictive Value of New Measurements (forum thread)</p><p>Thoughts: Forums can be great for asking the author for exact answers to complex questions</p><p><a href="https://mastodon.social/tags/modelselection" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>modelselection</span></a> <a href="https://mastodon.social/tags/causalinference" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>causalinference</span></a> <a href="https://mastodon.social/tags/prediction" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>prediction</span></a> <a href="https://mastodon.social/tags/bias" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>bias</span></a> <a href="https://mastodon.social/tags/information" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>information</span></a></p><p><a href="https://discourse.datamethods.org/t/statistically-efficient-ways-to-quantify-added-predictive-value-of-new-measurements/2013/1" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">discourse.datamethods.org/t/st</span><span class="invisible">atistically-efficient-ways-to-quantify-added-predictive-value-of-new-measurements/2013/1</span></a></p>
Dr Mircea Zloteanu ☀️ 🌊🌴<p><a href="https://mastodon.social/tags/statstab" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>statstab</span></a> #358 What are some of the problems with stepwise regression?</p><p>Thoughts: Model selection is not an easy task, but maybe don't naively try step wise reg.</p><p><a href="https://mastodon.social/tags/stepwise" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>stepwise</span></a> <a href="https://mastodon.social/tags/regression" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>regression</span></a> <a href="https://mastodon.social/tags/QRPs" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>QRPs</span></a> <a href="https://mastodon.social/tags/issues" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>issues</span></a> <a href="https://mastodon.social/tags/phacking" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>phacking</span></a> <a href="https://mastodon.social/tags/modelselection" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>modelselection</span></a> <a href="https://mastodon.social/tags/bias" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>bias</span></a></p><p><a href="https://www.stata.com/support/faqs/statistics/stepwise-regression-problems/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">stata.com/support/faqs/statist</span><span class="invisible">ics/stepwise-regression-problems/</span></a></p>
SRF IRIS<p>IRIS Insights I Nico Formanek: Are hyperparameters vibes?<br>April 24, 2025, 2:00 p.m. (CEST)<br>Our second IRIS Insights talk will take place with Nico Formanek.<br>🟦 <br>This talk will discuss the role of hyperparameters in optimization methods for model selection (currently often called ML) from a philosophy of science point of view. Special consideration is given to the question of whether there can be principled ways to fix hyperparameters in a maximally agnostic setting.<br>🟦 <br>This is a WebEx talk to which everyone who is interested is cordially invited. It will take place in English. Our IRIS speaker, Jun.-Prof. Dr. Maria Wirzberger, will moderate it. Following Nico Formanek's presentation, there will be an opportunity to ask questions. We look forward to active participation.<br>🟦 <br>Please join this Webex talk using the following link:<br><a href="https://lnkd.in/eJNiUQKV" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="">lnkd.in/eJNiUQKV</span><span class="invisible"></span></a><br>🟦 <br><a href="https://xn--baw-joa.social/tags/Hyperparameters" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Hyperparameters</span></a> <a href="https://xn--baw-joa.social/tags/ModelSelection" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ModelSelection</span></a> <a href="https://xn--baw-joa.social/tags/Optimization" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Optimization</span></a> <a href="https://xn--baw-joa.social/tags/MLMethods" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>MLMethods</span></a> <a href="https://xn--baw-joa.social/tags/PhilosophyOfScience" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>PhilosophyOfScience</span></a> <a href="https://xn--baw-joa.social/tags/ScientificMethod" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ScientificMethod</span></a> <a href="https://xn--baw-joa.social/tags/AgnosticLearning" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AgnosticLearning</span></a> <a href="https://xn--baw-joa.social/tags/MachineLearning" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>MachineLearning</span></a> <a href="https://xn--baw-joa.social/tags/InterdisciplinaryResearch" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>InterdisciplinaryResearch</span></a> <a href="https://xn--baw-joa.social/tags/AIandPhilosophy" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AIandPhilosophy</span></a> <a href="https://xn--baw-joa.social/tags/EthicsInAI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>EthicsInAI</span></a> <a href="https://xn--baw-joa.social/tags/ResponsibleAI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ResponsibleAI</span></a> <a href="https://xn--baw-joa.social/tags/AITheory" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AITheory</span></a> <a href="https://xn--baw-joa.social/tags/WebTalk" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>WebTalk</span></a> <a href="https://xn--baw-joa.social/tags/OnlineLecture" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>OnlineLecture</span></a> <a href="https://xn--baw-joa.social/tags/ResearchTalk" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ResearchTalk</span></a> <a href="https://xn--baw-joa.social/tags/ScienceEvents" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ScienceEvents</span></a> <a href="https://xn--baw-joa.social/tags/OpenInvitation" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>OpenInvitation</span></a> <a href="https://xn--baw-joa.social/tags/AICommunity" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AICommunity</span></a> <a href="https://xn--baw-joa.social/tags/LinkedInScience" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LinkedInScience</span></a> <a href="https://xn--baw-joa.social/tags/TechPhilosophy" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>TechPhilosophy</span></a> <a href="https://xn--baw-joa.social/tags/AIConversations" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AIConversations</span></a></p>
Nicola Romanò<p>Can anyone help with understanding how to best do <a href="https://qoto.org/tags/modelselection" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>modelselection</span></a> in the context of <a href="https://qoto.org/tags/neuralnetworks" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>neuralnetworks</span></a> ? I'm trying to understand how to reduce <a href="https://qoto.org/tags/bias" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>bias</span></a> due to the selection of a particular test set.</p><p>More details here</p><p><a href="https://stats.stackexchange.com/q/620547/582" rel="nofollow noopener" target="_blank"><span class="invisible">https://</span><span class="ellipsis">stats.stackexchange.com/q/6205</span><span class="invisible">47/582</span></a></p>
katch wreck<p>from the standpoint of model selection, parsimony often boils down to dimensionality reduction</p><p><a href="https://mastodon.social/tags/modelSelection" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>modelSelection</span></a> <a href="https://mastodon.social/tags/parsimony" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>parsimony</span></a> <a href="https://mastodon.social/tags/OccamsRazor" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>OccamsRazor</span></a> <a href="https://mastodon.social/tags/dimensionalityReduction" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>dimensionalityReduction</span></a> <a href="https://mastodon.social/tags/degreesOfFreedom" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>degreesOfFreedom</span></a> <a href="https://mastodon.social/tags/complexity" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>complexity</span></a> <a href="https://mastodon.social/tags/informationTheory" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>informationTheory</span></a> <a href="https://mastodon.social/tags/biasVarianceTradeoff" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>biasVarianceTradeoff</span></a> <a href="https://mastodon.social/tags/overfitting" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>overfitting</span></a> <a href="https://mastodon.social/tags/underfitting" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>underfitting</span></a> <a href="https://mastodon.social/tags/optimization" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>optimization</span></a> <a href="https://mastodon.social/tags/parameterTuning" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>parameterTuning</span></a> <a href="https://mastodon.social/tags/crossValidation" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>crossValidation</span></a> <a href="https://mastodon.social/tags/inverseProblems" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>inverseProblems</span></a> <a href="https://mastodon.social/tags/inference" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>inference</span></a> <a href="https://mastodon.social/tags/statisticalLearning" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>statisticalLearning</span></a> <a href="https://mastodon.social/tags/machineLearning" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>machineLearning</span></a> <a href="https://mastodon.social/tags/ML" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ML</span></a> <a href="https://mastodon.social/tags/dataScience" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>dataScience</span></a> <a href="https://mastodon.social/tags/modeling" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>modeling</span></a> <a href="https://mastodon.social/tags/decisionTheory" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>decisionTheory</span></a> <a href="https://mastodon.social/tags/fitting" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>fitting</span></a> <a href="https://mastodon.social/tags/regression" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>regression</span></a> <a href="https://mastodon.social/tags/classification" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>classification</span></a> <a href="https://mastodon.social/tags/residualError" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>residualError</span></a> <a href="https://mastodon.social/tags/costFunction" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>costFunction</span></a> <a href="https://mastodon.social/tags/performanceLoss" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>performanceLoss</span></a></p>
Blake Richards<p>7/10) This finding led to our <a href="https://fediscience.org/tags/proposal" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>proposal</span></a>: Can we use α for <a href="https://fediscience.org/tags/modelSelection" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>modelSelection</span></a> in an <a href="https://fediscience.org/tags/SSL" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>SSL</span></a> pipeline? </p><p>Two key +s of α:</p><p>1. α doesn’t require labels</p><p>2. α is quick to <a href="https://fediscience.org/tags/compute" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>compute</span></a> (compared to training a readout)</p><p>We study hyperparam selection in <a href="https://fediscience.org/tags/BarlowTwins" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>BarlowTwins</span></a> (Zbontar et al.) as a case study!</p><p><a href="https://fediscience.org/tags/AI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AI</span></a> <a href="https://fediscience.org/tags/ML" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ML</span></a> <a href="https://fediscience.org/tags/deeplearning" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>deeplearning</span></a> <a href="https://fediscience.org/tags/neuroscience" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>neuroscience</span></a></p>