Singular Value Decomposition (SVD): Overview
HTML-код
- Опубликовано: 22 авг 2024
- This video presents an overview of the singular value decomposition (SVD), which is one of the most widely used algorithms for data processing, reduced-order modeling, and high-dimensional statistics.
These lectures follow Chapter 1 from:
"Data-Driven Science and Engineering: Machine Learning, Dynamical Systems, and Control" by Brunton and Kutz
Amazon: www.amazon.com...
Book Website: databookuw.com
Chapters available at: databookuw.com/...
Brunton Website: eigensteve.com
This video was produced at the University of Washington
Oh my god, was going through the control bootcamp, hit a wall with SVD, the next day, series in SVD. Thanks a lot for these gorgeous lectures.
"I'm gonna walk you through..."
Dude has no legs and is still gonna walk. #Inspirational
how do I like this twice
He also seems to be able to apply a mirroring transformation allowing him to write backwards.
I can't unsee that now lmao
and write words in mirrow form omg
This is a fantastic channel Dr. Brunton! I just stumbled across it searching for SVD and got hooked on the content. Your unpretentious, clear and thorough explanations are a delight. They way you've set up the board and screen projection is fantastic. I'm really impressed with this and want to say thank you to you and the team that made this production possible.
It's really interesting to see how left-handedess works when mirrored. Quite an effect on the brain!
Dude, I've been trying to figure out whether he's writing everything backwards, which side his parting is on, whether he's left-handed or right-handed, married or not...
Do you understand what magic is this?!
Haha... no worries - I've figured it out!
@dan do tell because I have no idea how this video was made
@@aishaaijazahmad6184 he writes normally on a piece of glass between him and the camera. Then, in editing, he mirrors the video
It looks like he’s writing backwards with his right hand. In reality, he’s writing normally with his left hand
@@beastmode1647 😂I was wondering how too ! Thank u
Truly a remarkable series of lectures. Infinitely appreciative.
Glad you like them!
hey, steve! just wanted to say that I am doing data science in college right now and it is really hard to learn when you don't understand why you are doing something but you make it really easy to understand why we do a certain thing and how it actually connects to data science. so thank you sm!
Hey I just came across your comment and was curious. How have your studies in data science been going?
is he filming through a mirror or writing backwards? or mirroring the video before posting? what is going on here, how is he writing backwards!?!
it's called a lightboard, just google it
"just google it". I guess some people can't help but be assholes
@@fixed-point did i hurt your feelings little man? I think you need to grow up. I don't know much except that it is called a lightboard and google definitely knows more than me. It is an admission that I'm not an expert. Who's the asshole, asshole
@@charlieolivergomez2376 looks like I hurt your feelings. I'm sorry 😢
@@fixed-point No offense taken brother. It was my fault in the first place for not being clear
Wow ! Wasn’t expecting a new video, especially on that topic though !
Great to see you back on lecturing! The new video aspect looks great with the black background.
Thank you !
These videos are priceless. Thank you for such a clear exposition of an important topic.
I love finding wonderful teachers unexpectedly
Discovering this channel is one of the best things that happened to me. Thank you isn't enough to say how I feel. wow
Really amazing how you introduce us to “mysterious” ways of manipulating our interests in such a crystal clear way!
Watched a lecture on SVD as part of a vector semantics course. Will watch the 2nd part and exercise tomorrow, but I added a few of your videos to my watch later and will watch them over night. Your rather visual explanations always help me to understand the concepts and see their use.
Awesome videos - thanks so much for the easy-to-understand explanations. I just ordered the book and this greatly helps supplement!
Machine Learning for Manufacturing Automation is a brand new course being offered at Missouri University of Science and Technology, and I am in the second class ever to go through it. Thanks for your videos, they’re very helpful to establish intuition!!
Thank you so much for the effort you put in to help others. So grateful for the quality content. I'm very privileged to have access to this quality of teaching!
Thank you, Dr. Brunton! Really hope you can upload more videos!
Great video! My notes: 2:32 3:15 3:26 used to solve non-square that are non-invertible
3:38 3:52 3:58 4:40 4:52 5:07 5:26
I am learning this because I saw this being used in Battery modelling for electric vehicles. I don't understand how it is used (which is why I am here) but you can add that to number of applications of SVD
Thanks!
never realized that Prof Kutz was your colleague. he really is a legend.
Excellent! Can you add numbers (in the series) so we know what order to watch them? Thanks
I have put everything into playlists, so that should help with the ordering.
Thank you! It is nice to have you back 😉
A big thanks to you Dr. Brunton !
This series helped me immensely. Thanks a lot!
Thank you so much for this video series. It helped me a lot!
SVD overview? More like "Super great information that everyone should view!" 👍
Wow soo good to have you back 😃....Thank you
Best lectures of SVD! BTW, Dr.Brunton looks like Filthy Frank (but more handsome than Papa Franku)
I have to say: not only the lectures are really handsome, the man himself is also really handsome. hahaha. What a great educator!
love it, what a series, thank
Super curious about how he writes in that direction with out hand writing flipped.😲
Very very excited for the series. Thank you
Awesome lecture!!! Writing in mirror style and still be able to have great flow of thought, woow
I always wondered how these videos are recorded. Is there a glass between you and the camera? Great content anyways!
Blind guess: Its glass and than they flip the video horizontally... Or Mr Brunton is not only a genius but is also able to write mirror font.
I finde both equally likely
@@Sonnentau1 Dr Brunton seems to be actually left-handed as per other videos; so the glass/flip-video hypothesis/trick seems likely. Neat.
Wedding band on the "right" hand, so think he's left-handed and image flipped :)
I have asked Prof brunton in another video about the exact same thing and he replied with 'years of practice' :)
If there is glass, the writing is still backwards relative to the writer. I don't see how it matters if you flip it.
You like like "Harrison wells" character in the flash season 1. He was a great scientist and teacher in Reel life , You are same in real life.
Very nice Videos, Thank you
Is there an order to all these videos? Difficult to discern. Except for the ones that begin, “Welcome back ...”.
Here is the playlist: ruclips.net/video/gXbThCXjZFM/видео.html
Thank you! I was just studying Word2Vec and GloVe algorithms and really needed this kind of content!
Just surprised at the production logic behind the video. Had to think for solid 5 min to figure out how it was recorded and edited for us to be able to read those words.
I got inner peace after figuring that out.
Thank you so much for these videos I'm persuing the playlist and I really love it!
i really like the video presentation and clarity of topics, stayed and subscribed.
The writing is trippy man. The words are not supposed to look like that if he’s facing the camera, it’s supposed to be from right to left and written backward. How does the writing looks like normal in the video? Unless the guy is writing backward
Concise and rich. Brilliant!
Very, very nice explanation and presentation. Thank you!
I tried to get a tailored basis for my data using SVD, and it yielded the exact Fourier basis. :D
Thanks so much for this series, I’m so excited to watch it! Just wondering if there are videos for all chapters of the book or only a select few?
Thanks for posting these videos
I would have changed the series name, as you have said @ 5:33, to "Linear Algebra to make Money" :-). Thank you Professor for sharing the series of lectures - very thorough and extremely helpful.
Hope the series could link to the colab, thanks for the content.
are you writing backwards on a glass board or is this some kind of editing magic ... this keeps me up at night
Thanks for the content! Also super impressive if it is the case that you're writing in reverse!
Glad you like it!
Thank you so much for the content.
For the first 3 minutes and I was so confused and amazed by his incredible talent for mirror writing (it looked as good as my normal handwriting!) Then I realized he probably flipped the video 😅😆
Amazing series!
Great stuff; thanks a lot. Quick question! Is it fair to say that the left singular vectors reflect the distribution (statistical distribution) of the column vectors?!
Great lectures. I have ordered your book.
are you left or right handed?
I did high graduate school in engineering and I learned there how to hate mathematics. If only I had such a teacher, it would have been definitely different.
Hey Steve. Silly question... Did you learn to write backward so you can do these videos or how do you achieve this? 😅
mirror the video ;)
@@and_jannis no it is not. He is actually writing from his right to left as backwards
ruclips.net/video/wCOuu0-o5YI/видео.html HE is probably shooting through a mirror
I love this, thank you!
Loved it content, presentation 💗💗💗
Fantastic presentation. I only know what the SVD was, had no idea it had so many uses.
I wish the com section was a bit more specific and bring ideas other than ' how to say Thank you ' to a man who probably knows how good he does, common guys, don't be the FFT, try to be the SVD and tailor your comments to this particular video..
Can not access some of the function since they are in the Statistical Toolbox.
You are left-handed. Gotcha!
Hi, Steve! Nice lecture. Is there a exercise for this topic?
Why this video is available only in HD? It can be a bit problematic...
i have all the options from 144p up to 1080p...
Strange. I still have only 1440p and 2160p options.
Sorry about that. Newer videos will have this issue fixed.
Which classes would be prerequisites for this content if it were to be presented in a class at a University?
Thanks!
Welcome!
If I have a single time series.. Then what should i do to get eigenvalues?
Thank you ))
Sir can you tell us the most important math lessons a student of automatic control and signals needs?
😊😊😊😊
Great video.
Wish my ADHD brain would stop focusing so much on the handwaving and the hows and why of the this weird otherworldly transparent glassboard/whiteboard
Steve, thanks for sharing!
Can you please clarify the how can combine results two or more singular value decomposition got from same object like multilayer matrix?
Hi, is there any sequence I should follow to watch these videos ?
What do that mean "Data-driven" generalization of FFT?
Where can I find -in the playlist- the "tailored" data-driven transforms part?
SVD is gangsta until Matrix: 1000000 x1000000 arrives!!
Then hit it with randomized SVD!
Simple, linear = Predictable or syntropy
The time domain is dual to the frequency domain
Thesis (time) is dual to anti-thesis (frequency) -- the time independent Hegelian dialectic
Predictability is dual to unpredictability, certainty is dual to uncertainty, syntropy is dual to increasing entropy, the fourth law of thermodynamics. Energy is dual to mass -- Einstein.
Dark energy is dual to dark matter.
what
@@dankons6578 There is a 4th law of thermodynamics, syntropy is a dual process to that of increasing entropy.
Teleological physics (syntropy) is dual to non-teleological physics (entropy).
Certainty is dual to uncertainty -- the Heisenberg certainty/uncertainty principle.
The conservation of duality is the 5th law of thermodynamics.
Duality is energy, energy is duality.
Potential energy is dual to kinetic energy.
Apples fall to the ground because they are conserving duality.
Action is dual to reaction -- Sir Isaac Newton.
Positive charge is dual to negative charge -- electric fields.
North poles are dual to south poles -- magnetic fields.
Electro is dual to magnetic, electro magnetic energy (light) is dual.
"Always two there are" -- Yoda.
oh ok i get it now
@@dankons6578 Gravitation is equivalent or dual to acceleration -- Einstein's happiest thought, the principle of equivalence (duality).
There is a pattern of duality hardwired into the physics and also philosophy, mathematics, biology, control theory etc.
This important fact is currently not appreciated by those who fund science.
i understand clearly how to compute the SVD. but it seems like that i forgot something. if i have a matrix of nxn and i want to reduce the dimesion to rxr . how do i get the rxr matrix? if someone knows inform me i need it right away.
Good Presentation. 👩🏽🏫. There is a book and .pdf listed above. Where can I find how correlated or not these documents are? Do these two documents work together? Does the .pdf document have an errata sheet? Thanks.
How does he write backwards?
I've done watched the videos, till data visualization - buyer beware. Please tell me, should I watch another one playlist before watch this video? Like linear algebra or begining scientific computing? Thank you
Is there some video for each chapter of the book?
Yes, we tried to break it down so each section has a video.
What application did you use to make this transparent board?
We are using the lightboard
Svd like when I be pulling out this Russian svd dragunov in the school
D - Decomposition
R - Rational
A - Angle
G - Geometry
U - Unitary
N - Numbers
O - Obstacle
V - Variation
Does someone now what marker pen this guys use ?
Guess I'm late to the party. Just found out your channel. Damnnnnn
Love from india❤️
Hello from Seattle!
Yo, how do you learn to write in reverse like that
I’m searching SVD gun and this video shows up
What dry erase markers do you use?
Hello Prof. I want to ask you something, is SVD have big impact on system identification of a dynamic system, if we combine with least square algorithm ? or it just useful enough for machine learning task ? Thanks you
at 5:34... How we can make money from Linear Algebra?
0:52 ! how it is possible? writing backward!!!
Hi! Thank you so much for these vids. Realized yesterday that I couldn't make myself use SVD unless I understand the logic of it.
Could someone possibly recommend the topics I should look into beforehand in order to understand it and subsequent topics? I don't have a foundation in Maths beyond High School Level.
you should learn some linear algebra. you need a little calculus to do linear algebra sometimes, but it can mostly be done with integration and differentation. easy topics to look into, you're going to be using basic integration and differentiation. linear systems, manipulating them, understand matrices and how to manipulate them, how to find eigen values and eigenvectors, what eigenvalues and eigenvectors are. kinda gives you a base level for linear algebra.
@@ronronn3148 Thank you so much! I think I know what all you've mentioned but I'm coming to realize that what I know generally isn't everything I'm supposed to know. I'll look up the mentioned topics in detail. :-D
@@all_bright There's alot of stuff you're gonna be lost on and I think its important to write stuff down, or take mental note of what vocabulary and operations makes your mind blank. You look those words up, read about the topics, and fill in the blanks. alot of resources online.
@@ronronn3148 Yes! Thank you for the suggestion. Will keep that in mind.
You don't even need to take notes. Just take a screenshot of the sketches
Am I stupid or does he have to be writing backwards for this to work?
These days if you say the words "Cloud" and "Machine Learning" in the same sentence venture capitalists will swarm
Today, I have discovered a gold mine...
nice explain how to interact with you
harrison wells