#jmlr — Public Fediverse posts

weekend yt lecture suggestion:

“What's wrong with #LLMs
&
what we should be building instead”

ಠ_ಠ

issues:
• Incorrect & contradictory answers
• Dangerous & socially unacceptable answers
• Expensive to train & lack of updatability
• Lack of attribution & poor non-linguistic knowledge

challenges:
• Sepparate Linguistics from World Knowdledge
• Functioning Episodic Memory

By Veteran since #GOFAI times, #MLPioneer #JMLR Journo etc:
Prof #ThomasGDietterich @tdietterich https://youtu.be/cEyHsMzbZBs

This paper considers the problem of choosing "good" #MachineLearning algorithms from their performance on a number of datasets. They use an idea from #psychometric testing called "item-response theory". A particular benchmark is a single question in a test (some datasets are hard, others are easy). A particular ML algorithm is a student taking a test (some algorithms are consistent, some do well on easy questions, others give puzzling performance).

#JMLR
https://jmlr.org/papers/v24/20-1318.html

'Sufficient reductions in regression with mixed predictors', by Efstathia Bura, Liliana Forzani, Rodrigo Garcia Arancibia, Pamela Llop, Diego Tomassi.

http://jmlr.org/papers/v23/21-0175.html

#NewPaper #JMLR

'Towards An Efficient Approach for the Nonconvex lp Ball Projection: Algorithm and Analysis', by Xiangyu Yang, Jiashan Wang, Hao Wang.

http://jmlr.org/papers/v23/21-0133.html

#NewPaper #JMLR

'Total Stability of SVMs and Localized SVMs', by Hannes Köhler, Andreas Christmann.

http://jmlr.org/papers/v23/21-0129.html

#NewPaper #JMLR

'Distributed Learning of Finite Gaussian Mixtures', by Qiong Zhang, Jiahua Chen.

http://jmlr.org/papers/v23/21-0093.html

#NewPaper #JMLR

'PECOS: Prediction for Enormous and Correlated Output Spaces', by Hsiang-Fu Yu, Kai Zhong, Jiong Zhang, Wei-Cheng Chang, Inderjit S. Dhillon.

http://jmlr.org/papers/v23/21-0085.html

#NewPaper #JMLR

'Unlabeled Data Help in Graph-Based Semi-Supervised Learning: A Bayesian Nonparametrics Perspective', by Daniel Sanz-Alonso, Ruiyi Yang.

http://jmlr.org/papers/v23/21-0084.html

#NewPaper #JMLR

'Rethinking Nonlinear Instrumental Variable Models through Prediction Validity', by Chunxiao Li, Cynthia Rudin, Tyler H. McCormick.

http://jmlr.org/papers/v23/21-0082.html

#NewPaper #JMLR

'Attraction-Repulsion Spectrum in Neighbor Embeddings', by Jan Niklas Böhm, Philipp Berens, Dmitry Kobak.

http://jmlr.org/papers/v23/21-0055.html

#NewPaper #JMLR

Along with the accepted papers at #NeurIPS2022, 18 posters from ML #Reproducibility Challenge (#MLRC2021) and 31 posters from #JMLR will be presented in the in-person session at the inaugural Journal Showcase Track at #NOLA the coming week! Posters will be presented in all the six in-person sessions: https://neurips.cc/virtual/2022/events/journal_track. Do drop by and say hi! 👋

Journal Showcase Track blogpost: https://blog.neurips.cc/2022/08/15/journal-showcase/

@djoerd
Yes! I had heard that #JMLR has significantly less production costs... something like 6.5 USD per article

https://blogs.harvard.edu/pamphlet/2012/03/06/an-efficient-journal/

Also see this discussion: https://ask-open-science.org/28/how-much-does-it-cost-to-publish-an-academic-article

P.S.: #JOSS (https://joss.theoj.org/) has net 0 costs by relying on GitHub for everything.