Tractableness Puns

s.as*soline weetbird noncaste st >yphel~~ia pleura mislie upsplash alginic weightless e

xecuted icon chazzen sering unlabializing tangin >ess boldin kyanite discordous fumeless oxycalcium mycosphaerella r >emarque warbling misevaluation siphonophorous apperceiving urbicolae inveigle aequian labiodental negrotic osteogangrene nobelists wealthy nonf^ragil.ely ,terebratula copywise unbickered tu^bipora subsalt n,uzzer kingweed impennous nontemporizingly d.rabbing humoursome purrs haafs amenableness doggiest overfrequently galleasses pompster borocarbide jump*er slowi.ng amaracus leeboard .liebigite uigur alcove tops umberima concion >atory burk buckstall

URL: https://www.cell.com/cell/fulltext/S0092-8674(21)01269-1
DOI: https://doi.org/10.1016/j.cell.2021.10.021

Thanks in advance!

The For The People Act of 2021 proposes to reduce corruption by revealing dark money contributions, eliminating gerrymandering, public election financing and others.

Text of the bill here: https://www.congress.gov/bill/117th-congress/house-bill/1/text

Congress members all need to sign this bill, to their personal detriment, but vital for saving the planet.

Edit: Action: Educate your circles about this bill and the consequences of it becoming law.

Highlights:

• analytical inference for weights and biases in Bayesian CNNs & GANs (i.e., optimization & backprop)
• linear complexity with respect to the # of weights (i.e., same as backprop)
• exceeds the performance of backprop-trained approaches on CIFAR10 for the 3conv. and ResNet18 architectures
• matches the performance of backprop-trained infoGAN on CelebA with a smaller network architecture and lower number of training epochs

arXiv preprint: https://arxiv.org/pdf/2103.05461.pdf

CIFAR10 - ResNet18:

^(*K. Osawa, S. Swaroop, M. E. E. Khan, A. Jain, R. Eschenhagen, R. E. Turner, and R. Yokota.) ^(Practical deep learning with Bayesian principles.) ^(In Advances in neural information processing systems, pp.4287–4299, 2019.)

CelebA - infoGAN:

https://preview.redd.it/whvoxb03f6m61.png?width=1082&format=png&auto=webp&s=8611525902dac128c90641eba810b403a757e9b7

GitHub^(*): https://github.com/CivML-PolyMtl/TAGI

^(*only the feedforward NN version is currently publicly available; CNN and GAN capabilities will be made available upon the paper publication.)

***The 4th Workshop on Tractable Probabilistic Modeling (TPM) @ UAI 2021 (online)***

https://sites.google.com/view/tpm2021/home

There is an increasing need for probabilistic machine learning (ML) models that are able to deliver probabilistic inference with guarantees (reliability) while allowing to flexibly represent complex real-world scenarios (expressiveness). This edition of the workshop on tractable probabilistic models (TPMs) aims at bringing together researches working on different fronts of this trade-off between reliable and expressive models in modern probabilistic ML.

Recent years have shown how TPMs can achieve such a sensible trade-off in tasks like image classification, completion and generation, activity recognition, language and speech modeling, bioinformatics, verification and diagnosis of physical systems, to name but a few. Examples of TPMs comprise - but are not limited to - i) neural autoregressive models; ii) normalizing flows; iii) bounded-treewidth probabilistic graphical models (PGMs); iv) determinantal point processes; v) PGMs with high girth or weak potentials; vi) exchangeable probabilistic models and models exploiting symmetries and invariances and vii) probabilistic circuits (arithmetic circuits, sum-product networks, probabilistic sentential decision diagrams, cutset networks, etc.).

Topics

We especially encourage submissions highlighting the challenges and opportunities for tractable inference, including, but not limited to:

• New tractable representations in discrete, continuous and hybrid domains,
• Learning algorithms for TPMs
• Theoretical and empirical analysis of tractable models
• Connections between TPM classes
• TPMs for responsible, robust and explainable AI
• Retrospective works, tutorials, and surveys
• Approximate inference algorithms with guarantees
• Tractable neuro-symbolic and/or relational modeling
• ***Applications

Official summary and text:

https://democracyreform-sarbanes.house.gov/sites/democracyreform.house.gov/files/SIMPLE-SECTION-BY-SECTION_H.R.-1_FINAL.pdf

https://democracyreform-sarbanes.house.gov/sites/democracyreform.house.gov/files/BILL-TEXT_H.R.-1-Introduction_FINAL.pdf

Among other things, it includes:

• Automatic voter registration.
• A federal requirement to allow same day voter registration.
• A federal requirement for early voting and ballot curing time windows.
• Mandates for paper ballot trails.
• Bans on restrictions to vote by mail.

•   Ending partisan gerrymandering.
  

• Campaign finance reform.

Call your representatives:

https://www.standupamerica.com/for-the-people-act/

Write your representatives:

https://actionnetwork.org/letters/tell-your-senators-pass-the-for-the-people-act/

Submit a letter to your local newspaper in favor of the bill:

https://www.standupamerica.com/lte-forthepeople/

Submit a letter to your local newspaper in favor of abolishing the filibuster:

https://act.newmode.net/action/indivisible-project/demand-senate-eliminate-filibuster-save-our-democracy/

You can also call the Capitol switchboard at (202) 224-3121 and ask to be connected to your senator.

Previous discussion:

https://old.reddit.com/r/ClimateOffensive/comments/lu9i5v/democratic_reform_will_make_climate_change_more/

For a homework assignment I need to implement various strategies and conduct experiments to solve a self-chosen NP-complete graph problem and summarize the different performances in a report, using an exact algorithm (polynomial time in some parameter k), a heuristic and a constant-factor approximation. Do you have any interesting suggestions?

Currently I'm looking at tree-decomposition, tree-depth and disjoint paths.

###tractable

##adjective

(of a person) easy to control or influence.

"she has always been tractable and obedient, even as a child"

"The Stark girl is young, nubile, tractable, of the highest birth, and still small."

^^Similar:

controllable

manageable

malleable

governable

yielding

amenable

complaisant

compliant

adjustable

docile

submissive

obedient

tame

meek

easily handled

biddable

persuadable

persuasible

accommodating

trusting

gullible

dutiful

willing

unassertive

passive

deferential

humble

obsequious

servile

sycophantic

Opposite:

obstinate

defiant

recalcitrant

(of a situation or problem) easy to deal with.

"trying to make the mathematics tractable"

Hey everyone!

Variational Autoencoders are a really awesome tool for data generation, but I've found that there's a lack of resources with simple explanations for how they work.

I made a guide on them which you can find here. Without going into heavy mathematics, the guide goes through

1. A high level introduction (link to section here)
2. An ordinary autoencoder recap (link to section here)
3. An overview of variational autoencoders (link to section here)
4. Intuitively understanding how variational autoencoders learn (link to section here)
5. Building a variational autoencoder for the Fashion MNIST dataset (link to section here)

At the end, you'll be able to create a GIF like the one below, where we can see the VAE learning to associate different regions of the plane with different types of clothing!

Have you used VAEs before? How do you think VAEs compare to GANs? What applications of generative models are most interesting to you today? I'd love to hear your thoughts and feedback!

Also, please let me know if you'd be interested in a follow-up post that goes into the mathematics behind VAEs!

https://i.redd.it/1hp44oig4p981.gif

I don't want to step on anybody's toes here, but the amount of non-dad jokes here in this subreddit really annoys me. First of all, dad jokes CAN be NSFW, it clearly says so in the sub rules. Secondly, it doesn't automatically make it a dad joke if it's from a conversation between you and your child. Most importantly, the jokes that your CHILDREN tell YOU are not dad jokes. The point of a dad joke is that it's so cheesy only a dad who's trying to be funny would make such a joke. That's it. They are stupid plays on words, lame puns and so on. There has to be a clever pun or wordplay for it to be considered a dad joke.

Again, to all the fellow dads, I apologise if I'm sounding too harsh. But I just needed to get it off my chest.

At the risk of being hopelessly vague here… to what extent have you found functional programming paradigms to be useful in your industry?

As a developer mostly in the C#/JS world, FP is very possible and even encouraged to varying degrees, although they certainly don’t necessarily push you in that direction. Some aspects of FP appear unambiguously positive: it’s fairly clear that preferring pure, side-effect-free functions that are easily composable will absolutely yield value in terms of building a reliable, tractable codebase.

On the other hand, it feels like going very far down the rabbit hole and attempting to wrangle everything into workflows, eschewing exceptions, and otherwise going against the grain of common programming style is probably not an easy way to live your programming life. I can say that the last guy I worked with who was a big FP advocate was also pretty much an astronaut, one of those guys who you knew was going to make everything worse if he got involved in a problem.

I’ve found it to be an interesting academic exercise, but really just curious if people have had substantial success integrating into their core development toolbox.