Heute um 17:30h ist die Finissage von https://techturk.form-f.art. Kommt vorbei: Lausitzer Str. 10, Sonnenhof.
Es gibt neues zu sehen + zu hören.
#techturk #foto #ml #machinelearing #algorithms #datasets #datamining

RT @: Great collection of relevant publications: Machine Learning in Proteomics and Metabolomics https://pubs.acs.org/doi/full/10.1021/acs.jproteome.2c00566?ref=vi_machine-learning-omics&utm_content=bufferd7719&utm_medium=social&utm_source=twitter.com&utm_campaign=buffer #MachineLearing #Massspectrometry #ArtificialIntelligence #Proteomics #Metabolomics #PrecisionMedicine #Biology #Biomarkers #Multiomics #Datasciences

Feels like combining #llm models with what Semantic Web is doing for a long time with triple stores and ontology modelling would be a good idea to enhance #ml #machinelearing #artificialinteligence #ai algorithms, has that been done already?

I got a new #machinelearing project! Yhehhh
But I don't know which models are perfect for my job.
I'm doomed :why:

Today's Large Language Models (LLMs, e.g. #ChatGPT or #Bard) are for the most part based on transformer technology, which is an attention-based approach (in fact, the paper that introduced transformer technology is called "Attention Is All You Need" [1]).

About a decade before the advent of transformers I wrote a bit about attention in neural nets here: https://davidmeyer.github.io/ml/attention.pdf. The LaTeX source is here: https://www.overleaf.com/read/gshqdkhqdnmm. My notes explain the basics of attention but are pretty old and incomplete (and obviously predate transformers). In any event, as always questions/comments/corrections/* are greatly appreciated.

#machinelearing #attention

References
--------------
[1] "Attention Is All You Need", https://arxiv.org/abs/1706.03762

🤔 How often do we really interrogate the drawings & figures in our DS & ML papers & educational content? Especially w.r.t. deep learning?

😬 I know I've had a bad habit of taking a "laissez-faire" approach to learning

🎉 I'm excited about this piece Aaron Master & Doron Bergman have written in support for more GOOD diagrams, especially in #deeplearning & #machinelearing papers.

https://medium.com/@amaster_37400/please-stop-drawing-neural-networks-wrong-ffd02b67ad77

This is a nice summary of attention and transformers in #machinelearing.

https://newsletter.theaiedge.io/p/the-aiedge-everything-you-need-to

My weekend reading, Deep Learning by John D.Kelleher, is short and to the point 👇🏼

#deeplearning #machinelearing

RT @EikoFried
1/4 If you read one #depression #biomarker paper this year, read this one by Nils & the gang. They looked at a large sample of depressed and healthy participants, investigating numerous features (neuro, genetics, etc) in 2.4 million #MachineLearing models. https://twitter.com/NilsRWinter/status/1633458191643312128

RT @EikoFried
1/4 If you read one #depression #biomarker paper this year, read this one by Nils & the gang. They looked at a large sample of depressed and healthy participants, investigating numerous features (neuro, genetics, etc) in 2.4 million #MachineLearing models. https://twitter.com/NilsRWinter/status/1633458191643312128

No one @eLife consulted me on the naming of this coll #machinelearing driven #computervision tool.

From: @eLife
https://fediscience.org/@eLife/109926639298527122

Daniel Jurafsky and James Martin's coursera class on Natural Language Processing is at the link. The class covers the foundations of modern search and language-oriented machine learning (AI) technologies.

https://youtu.be/CXpZnZM63Gg

#machinelearing #nlp #compsci #computerscience #datascience #search

Starting to realize what corporations are calling "AI" isn't what I think artificial intelligence is, but just statistics applied to a metric shit ton of stolen data.

#ai #statistics #machinelearing

Need to analyze a collection of datasets based on their #PersistenceDiagrams?
Check out our new approach for Principal Geodesic Analysis in the Wasserstein space of #PersistenceDiagrams, with applications to dimension reduction:
https://arxiv.org/abs/2207.10960
#TopologicalDataAnalysis #Visualization #DataScience #MachineLearing #TopologyToolKit
Funded by the European Research Council (ERC) (project TORI, https://erc-tori.github.io/)

This seems like a promising approach: a bit of physics informed and has flavor of DDA.
https://pubs.acs.org/doi/full/10.1021/acsphotonics.2c01019
'a Graph Neural Networks (GNN) architecture which learns to model #electromagnetic scattering, can be applied to metasurfaces of arbitrary sizes. Most importantly, it takes into account the coupling between scatterers. Using this approach, near-fields of #metasurfaces with dimensions spanning hundreds of times the wavelength can be obtained in seconds.' #photonics #GNN #MachineLearing

Teletext archaeology
Teletext is an information service accessible through televisions in Europe from the early 1970s onwards.
A team of archivists are now looking through old home videotapes to extract this data, unlocking these hidden incidental digital time capsules of old news listings which are mostly fully navigable…

https://m.facebook.com/story.php?story_fbid=pfbid0nPTvWLTMXN2xLW1BHdivjBuiauGYPe2tTuxEsFoEnnj8Wy98QkdbYgq4Bi8ZhEbsl&id=771203558
#technology #history #machinelearing

#aiart #machinelearing #stablediffusion #imgsynth
textual inversions, with poor mans outpainting

Hey #Seattle #machinelearing join me at Flatstick Pub on Nov 9 to learn about Spiking Neural Networks and neuromorphic computing 🤖 https://www.meetup.com/seattle-daml/events/289060384/

Hey #Seattle #machinelearing join me at Flatstick Pub on Nov 9 to learn about Spiking Neural Networks and neuromorphic computing 🤖 https://www.meetup.com/seattle-daml/events/289060384/

Is this thing on? Hi 👋 I'm James, I'm a 33 yo male (he/him) based in the UK. I'm a #STEM academic with interests in #machinelearing and #computervision, and a passion for #electricvehicles, #boardgames, #progmetal, and #whisky. Looking forward to see how Mastodon grows following the #twitterexodus!