These are public posts tagged with #ann. You can interact with them if you have an account anywhere in the fediverse.
ANN vs. SVM: When to Choose Artificial Neural Networks Over Support Vector Machines
ANNs are great for complex relationships but can be slow. SVMs are faster in high dimensions and less prone to errors with complex data. Which is better depends on your task and data. #ANN #SVM #MachineLearning #ArtificialIntelligence #DeepLearning #Comparison
https://tech-champion.com/artificial-intelligence/ann-vs-svm-when-to-choose-artificial-neural-networks-over-support-vector-machines/
Comparing ANN and...
Learn the advantages of artificial neural networks…
TECH CHAMPIONANN vs. SVM: When to Choose Artificial Neural Networks Over Support Vector Machines
ANNs are great for complex relationships but can be slow. SVMs are faster in high dimensions and less prone to errors with complex data. Which is better depends on your task and data. #ANN #SVM #MachineLearning #ArtificialIntelligence #DeepLearning #Comparison
https://tech-champion.com/artificial-intelligence/ann-vs-svm-when-to-choose-artificial-neural-networks-over-support-vector-machines/
Comparing ANN and...
Learn the advantages of artificial neural networks…
TECH CHAMPIONTrump's lawsuit against Ann Selzer is textbook #authoritarianism
Reading the complaint in Donald Trump’s lawsuit against
pollster #Ann #Selzer
over her 2024 poll that found Kamala Harris leading Trump in Iowa,
it’s hard not to give in to thoughts of how comically obscene it is.
The suit stretches the interpretation of the relevant law
— Iowa’s Consumer Fraud Act
— well past the breaking point.
It leans heavily on “facts” that are nothing but Trump quoting Trump about being mad.
It spends a good deal of time citing hard-right outlets like Breitbart as if they are neutral.
It’s hard not to laugh at how absurd it is.
-- But then you remember that ABC just gave Trump a total of $16 million rather than fight the paper-thin defamation lawsuit he brought against them.
And that even after ABC knuckled under, Trump still haslawsuits against CBS, Simon & Schuster, and CNN.
And then you also remember that Trump has stated he’s going to use all his might
— and that of the government
— to bury media he doesn’t like.
All of this makes trying to rationally assess whether the lawsuit will succeed nearly impossible.
Trump can afford to go scorched earth whether his claims have any merit or not.
He has bottomless wealth,
particularly regarding legal fees,
as he’s made his donors cover at least $100 million of them so far.
He’s utterly unconcerned about whether something is true,
and he seems to have a never-ending stream of lawyers willing to step up
even though several have ended up facing sanctions, criminal charges, or both.
Put simply, he has no incentive to back down, ever.
All that being said, the lawsuit against Selzer is still straight-up trash,
even if it’s trash that somehow manages to succeed because of the unique blend of horrible characteristics exclusive to Trump
https://www.publicnotice.co/p/trump-ann-selzer-lawsuit-explained?utm_campaign=post&utm_medium=web
In a sane country it'd be comical. But ...
www.publicnotice.coTrump’s retribution may focus on individuals,
but it’s a collective harm
-- So it makes sense to spread the cost of dealing with it.
In the wake of Donald Trump’s election victory and promised revenge tour,
a number of individuals have proposed the creation of an organization or fund
which would take on the job of defending the various lawsuits, prosecutions and generalized legal harassment Trump will bring to the table in the next four years.
It’s a very good idea.
It’s a necessary one.
So a few days ago I started reaching out to some people in the legal world and anti-Trump world to find out what’s going on,
whether any efforts are afoot and who is doing what.
What I found out is that there are at least a couple groups working toward doing something like this.
But the efforts seem embryonic.
Or at least I wasn’t able to find out too much.
The overnight news that Trump is now suing #Ann #Selzer and the #Des #Moines #Register over her final election poll for “election interference” makes me think that these efforts aren’t coming together soon enough
or can’t come together soon enough.
(If you’re not familiar with the details, Selzer is a pollster of almost legendary status
and in what turned out to be her final public poll, dramatically missed not only the result of the election but the whole direction of it.)
Let’s take the Selzer/Des Moines Register suit as our example.
Trump is claiming that he was damaged and should be made whole
because of a poll that showed him behind and turned out to be wrong.
His lawyers are trying to shoe-horn this claim into an Iowa consumer fraud statute.
But we shouldn’t be distracted by that.
The idea that a political candidate has a cause of action over a poll is absurd on its face.
And really that is precisely the point.
Trump casts penumbras of power and fear with talk; he holds public space; he keeps opponents off balance and guessing.
This is another example.
A lot of the power and point of such an exercise is precisely the absurdity of it.
It is meant to spur a chorus of “You can’t do that” and
“How can he do that?”
But he does do it.
We have that same mixture of outrage, incomprehension, uncanny laughter,
the upshot of which is an overwhelming and over-powering belief that the rules somehow don’t apply to this guy.
That’s his power and that is the point.
It is a performance art of power enabled by a shameless abuse of the legal system
Trump’s opponents need to learn to speak in that language.
Otherwise it’s a professional wrestling grudge match,
a taunt-fest with only one side taunting.
I fear some folks simply don’t get this dimension of what’s going on.
It’s not something you learn in law school and not in conventional, old-style politics either.
Defending the targets is key. But that’s not the only point or even the main one.
It’s about demonstrating the limits of Trump’s power and embarrassing him, President or not.
In the wake of Donald Trump’s election victory and…
TPM - Talking Points MemoA Machine Learning-Based Approach For Flash Flood Susceptibility Mapping Considering Rainfall Extremes In The Northeast Region Of Bangladesh
--
https://doi.org/10.1016/j.asr.2024.10.047 <-- shared paper
--
#GIS #spatial #mapping #water #hydrology #climatechange #extremeweather #flood #flooding #flashflood #maps #susceptibility #Bangladesh #remotesensing #field #historical #rainfall #extremes #AI #machinelearning #ANN #RNN #RFGB #CatBoost #risk #hazard #model #modeling #planning #management #preparation #prediction #precipitation
I just added some extra chapters on #ANN. Since we are using #autoencoders, I thought it could be useful to provide some general introduction on #NeuralNetworks and how they can be tuned.
A wild #ANN implementation in #jlang that I found on GitHub:
https://github.com/dimgeo/neural
Pretty cool stuff.
The nobel price for physics goes to Hinton and Hopfield for their work in what we now call AI.
https://www.nobelprize.org/uploads/2024/09/advanced-physicsprize2024.pdf
Backed by a cabal of wealthy conservative patrons like industrialist #David #Koch,
banker #Richard #Mellon #Scaife,
and the devout Catholic entrepreneur #Frank #Hanna,
the Federalist Society under Leo became a breeding ground for conservative judges who were recruited at law school,
groomed through the society’s program of events and talks,
and then bound together through their careers.
“The key was to figure out how to develop what I call a ‘pipeline’
— basically, where you recruit students in law school,
you get them through law school,
they come out of law school,
and then you find ways of continuing to involve them in legal policy,” Leo later explained.
In 2005, the Federalist Society began openly advocating for #John #Roberts
— a former member
— to be nominated to fill a vacant seat at the Supreme Court,
the first time it had campaigned publicly for a particular candidate.
A few months later, its sway had grown so much that it torpedoed President George W. Bush’s own preferred candidate for another vacant seat on the Supreme Court
— #Harriet #Miers, a judge and close friend of the president who wasn’t a member of the Federalist Society
— and pressured him to nominate #Samuel #Alito, one of its members, in her place.
Leo worked closely with the "Judicial Confirmation Network",
a new nonprofit organization set up using funds from #Robin #Arkley, a California businessman known as the
“foreclosure king,” who had made billions buying up mortgages of people in financial difficulties.
The idea for #JCN had been hatched at a dinner in Washington attended by Leo and Supreme Court Justice Antonin Scalia shortly after Bush’s reelection in late 2004.
JCN spent hundreds of thousands of dollars on radio and online advertisement to shape public opinion.
It was run by #Neil and #Ann #Corkery, a couple who had been members of #Opus #Dei since at least the eighties.
Neil had been a critical figure in getting a new residence for male, celibate members of the Catholic movement built in Reston, Virginia.
“Opus Dei members preach their faith through their work as well as the friendships they develop,” Ann explained.
She and her husband would later preach their faith by becoming central figures in a series of nonprofits that would channel dark money for Leo’s efforts.
Как найти иголку в стоге сена? Или обозор Retrieval Algorithms
Появление трансформеров, а впоследствии LLM (Large Language Models) привело к активному распространению чат-ботов и различных ассистентов помогающих в получении информации или генерации контента. Но несмотря на то что LLM способны по запросу генерировать человекоподобные тексты, они подвержены галлюцинациям . Естественным кажется желание уменьшить количество не достоверных ответов. Для этого мы можем либо дообучить LLM на наших данных, либо использовать Retrieval Augmented Generation (RAG) . RAG - это способ генерации текстов на новых данных без дообучения модели, с помощью добавления релевантных документов в промпт модели. Документы для генерации ищутся с помощью retrieval системы, после чего объединяются в один промпт и подаются в LLM для последующей обработки. В этой статье я решил собрать информацию о всех наиболее известных и применяемых алгоритмах поиска, с описаниями и материалами для более глубокого изучения.
https://habr.com/ru/articles/840268/
#bm25 #tfidf #hnsw #hnswlib #kd_tree #ann #product_quantization #LSH
Появление трансформеров, а впоследствии LLM (Large…
habr.comBacklash erupts over criticism of Tim Walz’s emotional son:
‘families are everything’
The intense emotional response of vice-president hopeful Tim Walz’s son to seeing his father on stage
– “That’s my dad!”
– thrust him into the spotlight as the convention’s breakout star
and with them raised the issue of nonverbal learning disorders, ADHD and an anxiety disorder that the 17-year-old is reported to experience.
The conservative commentator #Ann #Coulter posted an article on X about Gus’s reaction, captioning it: “Talk about weird …”
But the backlash against the mean-spirited post was swift.
“He’s 17” trended on the platform and Coulter – unusually – took the post down.
️“I can see why a child loving their parents would feel foreign to you,” wrote Obama staffer-turned-liberal podcaster Tommy Vietor.
It’s never been easy as the child of a high-profile US candidate, but the latest crop of presidential or vice-presidential hopeful offspring may be getting it worse than most,
and there is no avoiding the spotlight of both traditional media and social media.
https://www.theguardian.com/us-news/article/2024/aug/24/democratic-convention-tim-walz-son-emotions-criticized-backlash?CMP=Share_iOSApp_Other
The vice-presidential nominee’s son, Gus, was unrestrained…
www.theguardian.com#BASEsearch technology update:
We ripped out the #DDC classifier created by Mathias Lösch 12 years ago and replaced it with one created by Christoph Broschinski. It is using #annif, the automated subject indexing toolkit.
DDC classification is what allows users to browse our index by subject area: https://www.base-search.net/Browse/Dewey
Christoph found out that the decision tree based #Omikuji https://github.com/tomtung/omikuji outperformed #ANN-based https://fasttext.cc/ so we are luckily using explainable AI #XAI.
More than 340 mio. scientific documents from more than…
www.base-search.netДизайн встраиваемой базы данных для ANN запросов: MusyaDB
Я решил задизайнить встраиваемую базу данных . Это даст вам почувствовать вкус настоящего инженерного искусства. Статья получилась размером с небольшую книгу и разбивается на две больших статьи. В первой части мы поймем с чего вообще начинается дизайн таких систем, выберем алгоритмы и модель вычислений.
Логотип В моей прошлой статье из-за минималистичности…
habr.comHIRING: ML Engineering Manager / Mountain View, California, United States USD 204K+
#ANN #ComputerScience #Engineering #Lidar #MachineLearning #MLmodels #PhD #Python #Radar #Research
Applied Intuition is hiring for Full Time ML Engineering…
ai-jobs.netHas anyone got references (or ideas/recommendations) for how to perform #data #augmentation on #scrnaseq data (to use in training #ann)?
This is the only paper I could find, but maybe I am not searching for the right thing...
Identifying cellular identities (both novel and well-studied)…
www.ncbi.nlm.nih.gov[Перевод] Аннотации vs Документационные комментарии vs Маркерные интерфейсы
Прежде чем обсуждать, когда использовать аннотации вместо комментариев, важно сначала понять, что такое аннотации и для чего они используются в Java. Аннотации не заменяют комментарии к документации (у них есть и другие применения), но во многих случаях (особенно в тех случаях, когда нам нужны аналогичные структурированные комментарии) они предпочтительнее комментариев. Причина в том, что аннотации являются типами в Java, как и любой другой тип данных, определяемый пользователем. И, следовательно, их можно использовать после того, как они определены. Аннотации в Java на самом деле представляют собой всего лишь форму интерфейсов, и, следовательно, определение типа аннотации похоже на определение интерфейса. Ключевому слову «интерфейс» в определении аннотации предшествует символ «@». Типичный комментарий к документации — это то, что обычно появляется в начале любого определения класса. Он содержит имя автора, версию, дату, дату последнего изменения и т. д. Итак, пример комментария будет выглядеть примерно так:
Аннотация vs Документационный комментарий Прежде чем…
habr.comUseful new observational technique.
"The goal of the neural network setup is to learn temperature patterns across the United States and then identify whether climate change effects have exceeded the range of natural variability that has occurred in the past. This is called the timing of emergence (ToE), which is the first year that the effect has clearly appeared."
No plot spoiling. Results at #OpenAccess article.
#ClimateChange
#ClimateResearch
#ANN
https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023EF003981?af=R
Queen Bees review: Why, yes, elders rock https://oldaintdead.com/queen-bees-review-why-yes-elders-rock/ ##eldersrock, #Ann-Margret, #ChristopherLloyd, #ElizabethMitchell, #EllenBurstyn, #HarrisonPowell, #JamesCaan, #JaneCurtin, #LorettaDevine, #MatthewBarnes, #QueenBees
Queen Bees stars some lively elders living in a retirement…
Old Ain't Dead