Spent the past weeks trying to grasp backpropagation mathematically. Spent countless days on videos and textbooks. This is by far the most intuitive explanation. Just subbed. Thank you.
the EXACTLY video im looking for and 2:58 is the GEM, everybody is trying to explain backprop of each single weight of each layer w11, w12, w13, then you will drown, image there are100 layers, how to calculate w11, w12, ... w100, even a super computer will blow up. You really need to look at all weights in a single layer as a MATRIX, this is the magic and HACK of linear algebra, only by doing this, no matter how deep the NN is and how many nodes there is in a layer, you will not be afraid of it
at 12:18 at the bottom right corner i think the matrix with the xs should be a 1 x 3 matrix and not a 3x1 matrix, otherwise the matrix multiplication wont work
Spent the past weeks trying to grasp backpropagation mathematically. Spent countless days on videos and textbooks. This is by far the most intuitive explanation. Just subbed. Thank you.
I am glad this was helpful and thank you for subscribing.
hands down best video about backpropagation out there. subbed
Thanks for subscribing!
I couldn't really understand what a computational graph was until I watched this video. Thank you very much.
Thank you!
the EXACTLY video im looking for and 2:58 is the GEM, everybody is trying to explain backprop of each single weight of each layer w11, w12, w13, then you will drown, image there are100 layers, how to calculate w11, w12, ... w100, even a super computer will blow up. You really need to look at all weights in a single layer as a MATRIX, this is the magic and HACK of linear algebra, only by doing this, no matter how deep the NN is and how many nodes there is in a layer, you will not be afraid of it
the best session tha i can recommend to anyone. I studied computer science and i am not good at math and this was really a saver to coninue on.
Thank you again! This was incredibly simple to understand and exactly what I was looking for.
I cannot express how thankful I'm !
You can present complex information in simplified manner, many thanks!
The only video that explained me everything
Thank you so much!
at 12:18 at the bottom right corner i think the matrix with the xs should be a 1 x 3 matrix and not a 3x1 matrix, otherwise the matrix multiplication wont work
Both vectors x and w are 3x1. Note that we are using the transpose of W, which means that the result would be (1 x 3) x (3 x1), which is scalar.
very good thanks
Thank you!
I love you brother your teaching
thanks for meaningful describe
Thank you!
Very helpful
Can you explain about optimizer state
Thanks!! It was very useful
Thanks!
Nice session
Can you please do a video on different cross validation techniques in machine learning
Great suggestion, thanks!
What is phi with a1? Is it activation function.?
Great video
Thanks!
Do i need to understand all this to use neural networks? I'm scared shitless but my curiculuum says i have to learn it
You're the goat
nice video, birader aksanınızdan anladım
Thank you!