Video yükleniyor...

Video Yüklenemedi

Ana Sayfaya Dön

Data often lie on a low-dimensional manifold embedded in a high-dimensional space. But these manifolds are often highly non-linear, making linear dimensionality reduction methods like PCA insufficient. This has motivated the development of non-linear dimensionality reduction.

212,854 görüntüleme • 1 yıl önce •via X (Twitter)

43 Yorum

Alec Helbling profil fotoğrafı
Alec Helbling1 yıl önce

I put together this visual interactive blog explaining Isomap, a classic non-linear dimensionality reduction technique. Please check it out:

Alec Helbling profil fotoğrafı
Alec Helbling1 yıl önce

If you are interested in seeing the code (I haven't cleaned it up much 😅).

XIN Chen profil fotoğrafı
XIN Chen1 yıl önce

Really a nice illustration, How did you make this animation? Manim?

Alec Helbling profil fotoğrafı
Alec Helbling1 yıl önce

Thanks! This was with d3.js

Lin Gui profil fotoğrafı
Lin Gui1 yıl önce

May I include this in my teaching slides? Very comprehensive!

Alec Helbling profil fotoğrafı
Alec Helbling1 yıl önce

Thank you! Absolutely!

Ramón Horváth profil fotoğrafı
Ramón Horváth1 yıl önce

People using the term “highly non-linear” should be exiled from society.

Carl Friedrich Gauss profil fotoğrafı
Carl Friedrich Gauss1 yıl önce

Yeah there are manifold methods such as UMAP

Karan Jagtiani profil fotoğrafı
Karan Jagtiani1 yıl önce

Nice visualization. Captures how linear methods miss the twists in manifold shape. Did you experiment with other methods like t-SNE or UMAP on this?

Alec Helbling profil fotoğrafı
Alec Helbling1 yıl önce

Thanks! Perhaps in the future

Raymond Chua profil fotoğrafı
Raymond Chua1 yıl önce

@karanjagtiani04 Would also like to use a version of tsne and Umap!

Karan Jagtiani profil fotoğrafı
Karan Jagtiani1 yıl önce

@alec_helbling t-SNE and UMAP do catch those complex structures better, especially for high-dimensional data with non-linear relationships.

mina profil fotoğrafı
mina1 yıl önce

Non-linear dimensionality reduction??? U mean KPCA 🤣?

Bretta Gatesberg e/acc profil fotoğrafı
Bretta Gatesberg e/acc1 yıl önce

Exactly. The universe is an NLDR algorithm where gravity is the learned manifold's geometry and quantum mechanics is the aliasing error from the projection. Observation triggers collapse by de-aliasing the state.

AnglofuturistParty profil fotoğrafı
AnglofuturistParty1 yıl önce

What like kernel PCA ?

Node 7709 profil fotoğrafı
Node 77091 yıl önce

Non-linear spectral fit is very good.

unecht profil fotoğrafı
unecht1 yıl önce

Aaahh that was my PhD research and my retarded and insufficient advisor, as well as the commitee made me quit

soulblocks profil fotoğrafı
soulblocks1 yıl önce

Neat

deep Manifold profil fotoğrafı
deep Manifold1 yıl önce

** From High-Order Nonlinear Data to Deep Manifold **

Try TLGen App | Brandon Starship 🇺🇸 profil fotoğrafı
Try TLGen App | Brandon Starship 🇺🇸1 yıl önce

@grok how to apply this in ai training?

Garrett_OwlAcademy profil fotoğrafı
Garrett_OwlAcademy1 yıl önce

Mor - Model Object Reduction, is the group of techniques I started using along with lots of projected geometric expansion styles. I will say using a standardized universal 64 bit tensor hash to represent projected data, and having that projected data 'perspective' be a low-dimensional manifold really allows optimal compression, as long as you follow proper physics and geometry rules. So a quick example, 72 spheres pack optimally into a hexagonal polygon. If you were to wrap a 6 point node stream around a toroid, and divide it into 72 sections, you will get a projected 4-sphere that can convert completely to a square. This square can be a matrix that stores the 2D low dimensional array data. It's all about properly back-tracking the steps with the encoders / decoders.

Michael Grüner profil fotoğrafı
Michael Grüner1 yıl önce

been a long time since I heard about Isomaps!

Max David Gupta profil fotoğrafı
Max David Gupta1 yıl önce

super cool and much needed for a relatively un-discussed technique ! do you find people use isomap often enough in interpretability work ?

Pannous profil fotoğrafı
Pannous1 yıl önce

I was hoping to see it uncured onto a line

Alec Helbling profil fotoğrafı
Alec Helbling1 yıl önce

See the bottom of the blog haha.

Giacomo Bruzzo profil fotoğrafı
Giacomo Bruzzo1 yıl önce

This is a model for politics

César Padrón profil fotoğrafı
César Padrón9 ay önce

Nice! When you don't squeeze the spiral into that shape, PCA handles it perfectly. That doesn't happen with a symmetric spiral.

Saket Tiwari profil fotoğrafı
Saket Tiwari1 yıl önce

Great blog! I work on data geometry and I definitely learned a few new things from reading your blog.

Alec Helbling profil fotoğrafı
Alec Helbling1 yıl önce

Thanks! That means a lot

barney profil fotoğrafı
barney1 yıl önce

I’m absolutely terrible at maths but this post was written and presented so well I understood it super well!

Ditectrev profil fotoğrafı
Ditectrev1 yıl önce

Looks so good!👍👍

Tom Benadryl profil fotoğrafı
Tom Benadryl1 yıl önce

Easier said than done!

iDare e/acc profil fotoğrafı
iDare e/acc1 yıl önce

Add Spatiotemporal Multidimensional to this. I've made some similar visualizations to this one.

Omar profil fotoğrafı
Omar1 yıl önce

using nonlinear methods is like giving your analysis x-ray vision. most real-world data hides in weird shapes, written by the nonlinear tools.

Yoni profil fotoğrafı
Yoni1 yıl önce

very nice work

Alec Helbling profil fotoğrafı
Alec Helbling1 yıl önce

Thank you!

Clark H profil fotoğrafı
Clark H1 yıl önce

Open Qs: any way to get at ‘latent space’ elegantly? Or does it always need ‘Big Data’, and even w Big Data does it require tool specialized to that latent space (eg transformer in case of Machine Learning)? Eg What specialized tool might be required to see latent space of Big5?

joe profil fotoğrafı
joe1 yıl önce

Intuitive and soothing

Caelan Orien Solas profil fotoğrafı
Caelan Orien Solas1 yıl önce

3.87d maniford in 7d space

Sendao Haz profil fotoğrafı
Sendao Haz1 yıl önce

you will never stop those of us who are free to manipulate the system! lie in a two-dimensional manifold tonight!

Gunn profil fotoğrafı
Gunn1 yıl önce

Nicely written! I was gonna ask if you'd do the same for UMap but you already got that covered haha

Charlie Manson profil fotoğrafı
Charlie Manson1 yıl önce

How does it relate to UMAP?

Ido Klein עדו קליין profil fotoğrafı
Ido Klein עדו קליין1 yıl önce

@grok how to apply in gis - mainly in creating roads from a raster layer

Benzer Videolar