35. Finding Clusters in Graphs
Вставка
- Опубліковано 15 тра 2019
- MIT 18.065 Matrix Methods in Data Analysis, Signal Processing, and Machine Learning, Spring 2018
Instructor: Gilbert Strang
View the complete course: ocw.mit.edu/18-065S18
UA-cam Playlist: • MIT 18.065 Matrix Meth...
The topic of this lecture is clustering for graphs, meaning finding sets of 'related' vertices in graphs. The challenge is finding good algorithms to optimize cluster quality. Professor Strang reviews some possibilities.
License: Creative Commons BY-NC-SA
More information at ocw.mit.edu/terms
More courses at ocw.mit.edu
Such an amazing presence of mind and sharpness at 84 years old ... Respect to Prof Strang
A country that makes scientific knowledge public will always always be an oasis and an example to be followed.
Absolutely correct.
Pena que alguns países como o Brasil não sigam o exemplo de apoiar ciência e educação de qualidades para todos.
As Americans now live in a country where it is simply inadmissible to speak certain (important) scientific truths in public because of political correctness, this is exactly why their country is doomed.
Indeed, the astonishing truth of our current society is that a Chinese or a Russian student now works within a much freer academic environment than a student in the USA!
@@cnidariantide4207 Your way of looking at the world is extremely important, but, I do not see knowledge sharing by Russians or Chinese. As this is not clear to me, I cannot infer that the russian or chinese academic environment is free.
@@antoniosales3059 Thanks, I like your calm way of posing a counterpoint!
I do know what you mean-actually, while I was searching for information about graph algorithms, I found some nice videos from the Higher School of Economics in Moscow, and I thought it was quite fresh, quite unusual really to find quality information from a Russian university on the internet. To an English-speaking user of the web, it's almost as if they don't exist, while MIT, on the other hand, has an excellent, possibly unparalleled, tradition of making its teachings open to others on the internet.
What I said about the situation in China vs. America, well, I know that's not what you're actually talking about, but I must defend myself nonetheless. Now, looking at what I wrote up there, part of me wants to say "That was pure hyperbole!"-and then I'm reminded about how bad the situation in the US really is. I was going to post this article: www.aaup.org/article/academic-freedom-and-china to show how wrong I was, but then I read the piece's single comment from an academic who was born in China and teaches now in America: "[In China] I'm shocked how freely they speak on campus and off campus... [in the USA] the admins told me that the student and alumni I never met can say and do whatever they want, including ordering me to take down my syllabus... this is worse than what I've witnessed during the Cultural Revolution†". Really makes you think.
†Of course, while things loosened up in the years of Jiang Zemin and Hu Jintao, academic freedom seems to have become greatly restricted once again in the period of consolidation under Xi Jinping, which might be surprising even to Chinese academics who enjoyed the cooling-down after the Cultural Revolution (and it certainly makes it difficult for me to make any definite claims, having never been to China!)
I am in awe of Dr. Strang. All of his material is golden, and he such an obvious passion for teaching. Amazing communicator and academic.
The courses taught by professor Gilbert is just aaaawesome!!!
Professor Strang this lecture explains in great detail Clusters in Graphs and how Eigenvectors determine Clusters. The role of Spectral Clustering and it's start with the Laplacian Matrix. Thanks to MIT for all these fantastic lectures.
professor and MIT... the world stay better with this actions... thank you... thank you... Yes, he's a legend....
So much respect and admiration for Professor Strang
Thank you professor Gilbert!
What a great professor he is.
Perfect explanation， thanks
He is my best prof.
Dang this is an amazing lecture
"The job is to make some sense out of it." He is funny haha.
This is a great lecture omg
He is abolute legend !
all hail to open courses
I was kinda hoping that Prof. Strang started this lecture with his comment on "Does DL usually work?" from last lecture... Was that part edited out?
I wanted it too, and don't know why
@@junhuichi3200 Thirded, and allow me to propose a reason. Knowing the controversies and limitations makes you a better scholar :P
Professor Gilbert strang looks young, omg.:)
3:08 how do you take (a_i - x) if a_i and x are not adjacent nodes?
Did I miss anything hha, why is the class only 34 min and 29 min for the last one
very intriguing exposition to say the least , however , am concerned about the statistical validity of such ontological relationships and properties within the assumed structure of the graph knowledge ? your feedback is highly appreciated
The GOAT
So mostly neurotic system that connect with machine that go with x at y high or lower
Chose a algorithm the Coonect with 01 2 code remember code number
How old is the professor? It seems being 80
He was 83 when this lecture was being recorded.
Hello everybody,comment one