Sir, I have seen your linear algebra lectures, read your book on introduction to linear algebra and calculus and now this lecture series. I have learned a lot from you. I would like to express my deep gratitude towards you. Thank you for contributions.
Man, it's crazy. I watch these MIT lectures on biochemistry or computer science, and they make me feel dumb. Then, I come to math/stats/machine learning (my areas of study) and suddenly they seem palatable and even easy at times. Knowledge can be so deep and so specialized that there is just no way any one person can have enough time to learn all of these topics!
I'm so excited! I'm now at the tail end of a Master's program for Data Science in part due to Dr. Strang's linear algebra lectures. My future job heavily uses CNNs and I get to keep improving with Dr. Strang. I literally owe him my livelihood.
Great course! I really enjoy it :-) Especially the lectures about cyclic and toeplitz matrices are very interesting. Mostly because I came across toeplitz before and wondered that Lapack doesnt seem to have any special routines for block toeplitz and alike.( if I now remember correctly what I was searching for...) Seems like there are some structures behind these matrices such that calulations reduce quite signigicantly ... even for block-Toeplitz? And therefore only quite a few codes out there for block toeplitz? Talking about big to very big Ax=b systems (1 Mio unknowns on a normal laptop computer), inverse of Hessian matrices, fairly sparse, nearest neighbour interactions
DR. Strang, this is another fantastic lecture on the Convolution Rule and how it's applied in the real world. Fast Fourier Transform has play important role in convolution over time. Convolution is useful because of Fast Fourier Transform. Convolution is also very useful in Signal and Systems Theory, which is part of Electrical Engineering.
No, ImageNet is just a training data. It consists of a bunch of pairs of image and description in words of what's in the image. It has nothing to do with neural networks, other than the fact that CNNs are currently the best tools to do image classification, and thus to win ILSVRC challenges.
![[TH] FS vs BME - VCT Ascension Pacific - Lower Bracket Final](http://i.ytimg.com/vi/wnygqsR2Fig/mqdefault.jpg)
Full respect to Prof. Strang. So excited to see his insights to trendy technologies and he breaking down how it works.
Best lectures in the world.intuitive and advanced and very informative intelligent..there the best.THANKS EMERITUS GILBERT STRANG.
G) g) ggvi
Sir,
I have seen your linear algebra lectures, read your book on introduction to linear algebra and calculus and now this lecture series. I have learned a lot from you. I would like to express my deep gratitude towards you. Thank you for contributions.
Man, it's crazy. I watch these MIT lectures on biochemistry or computer science, and they make me feel dumb. Then, I come to math/stats/machine learning (my areas of study) and suddenly they seem palatable and even easy at times. Knowledge can be so deep and so specialized that there is just no way any one person can have enough time to learn all of these topics!
The fact that "I feel the very opposite for your first point" supports your second point...... my research field is about computer vision LoL
I'm so excited! I'm now at the tail end of a Master's program for Data Science in part due to Dr. Strang's linear algebra lectures. My future job heavily uses CNNs and I get to keep improving with Dr. Strang. I literally owe him my livelihood.
Words would not suffice the knowledge that you have given us Dr. Strang
Thank you very much for these lectures.
Great course! I really enjoy it :-) Especially the lectures about cyclic and toeplitz matrices are very interesting. Mostly because I came across toeplitz before and wondered that Lapack doesnt seem to have any special routines for block toeplitz and alike.( if I now remember correctly what I was searching for...)
Seems like there are some structures behind these matrices such that calulations reduce quite signigicantly ... even for block-Toeplitz? And therefore only quite a few codes out there for block toeplitz? Talking about big to very big Ax=b systems (1 Mio unknowns on a normal laptop computer), inverse of Hessian matrices, fairly sparse, nearest neighbour interactions
Thank you very much for this interesting presentation.
Thank you SIR.
The last four minutes really confuse me. What is doing here? What is Laplacian and Why it comes to here?
What a big chalk!!!
Full respect to Prof. Strang.
good bless you, dear Professor
Lg
FCP
黑色素
DR. Strang, this is another fantastic lecture on the Convolution Rule and how it's applied in the real world. Fast Fourier Transform has play important role in convolution over time. Convolution is useful because of Fast Fourier Transform. Convolution is also very useful in Signal and Systems Theory, which is part of Electrical Engineering.
No, ImageNet is just a training data. It consists of a bunch of pairs of image and description in words of what's in the image. It has nothing to do with neural networks, other than the fact that CNNs are currently the best tools to do image classification, and thus to win ILSVRC challenges.
It's clear you didn't follow the course , you've just read the title and replied
I read the title and the caption. Enough to make me leave.
Yes, the title is not good. But the lecture is useful if you don't know convolution!
@@mohamedberrimi4850 so what is ImageNet i am confused here
@@mohamedberrimi4850 I am clearly have not watch the entire lecture