The most clear explanation I found on youtube. Thank you so much. Please make more concept videos like this about machine learning

Nice explanation. Actually, I have seen all theory till now, you showed how backpropagation is actually calculating the further weights and biases. Thanks again.

This might seem like nothing, but just wanted to say that I did enjoy the video, made back propagation easy to understand without superficial explanation. Going deep right into the heart of the problem. Thanks alot man, was just what I wanted. One Love from Nigeria.

Simply the best of the best. Thank you for your hard work, thank you.

Finally got the idea about backprop..thanks man

Excellent explanation of the concept! Thank you so much for making this...

Thanks! This video explains back propagation very well! btw i almost passed out when the volume increased in the end

@CodeWrestling

3 жыл бұрын

Sorry for the volume.

Surely, this would help me a lot during my end semester exam. Thanks a lot 🙏

Excellent example!, Now I understand backpropagation and better for the steepest decent with the chain rule!

Can you remove the background music, otherwise it's awesome

@CodeWrestling

4 жыл бұрын

Cannot remove in this video, but will take care of it in coming videos.

@coxixx

3 жыл бұрын

@@CodeWrestling every rose has its thorn

@Ashajyothi23

3 жыл бұрын

I agree

@saisudha6512

2 жыл бұрын

very well explained

You are freaking best at explaining thank you

If now I can work in the field of data science, it's because of YOU 🙏

finally i understood back propagation.. Thank you so much!

@CodeWrestling

4 жыл бұрын

Thanks for appreciating!!

I wonder backpropagation was this much easy... thanks a ton 🙂🙂

This is the most clear explanation of back propagation I found . Thank you very much

@CodeWrestling

Жыл бұрын

Glad it was helpful!

What a Suplendid Explanation Love you Bro Stay Blessed

this site is just wonderful. Thank you so much for all the videos. It would be fantastic if you could explain their implementations in python also.

@dhanushp1680

2 жыл бұрын

this is not a site babe

Faar better explanation of Math than the other videos on KZread.

Thanks a ton, finally understood backpropagation (didn't know the math behind it was this easy)

Thank you so much bro for giving detailed explanation.

Thank you! You explained the crux of neural net in simple terms.

@CodeWrestling

4 жыл бұрын

Thanks a lot!!

Thank you, this video is thoroughly useful.

best one among all i have seen ever

Thanks a lot for such a beautiful explanation.

Great education video... Thanks!

You taught it very well to a backbencher like me!

@CodeWrestling

3 жыл бұрын

thanks, one of the best comment ever received.

You dont wanna mess with backpropagation ever never. Haha But he explain so good

very clear!! thank you!!!

@CodeWrestling

Жыл бұрын

Glad it helped!

Very good explanation . It cleared all my doubts.

Great video, thank you!

@CodeWrestling

4 жыл бұрын

Thanks for appreciating...!! #CodeWrestling

Thanks. It was an awesome explanation on the basics.It was very useful for me.

@CodeWrestling

3 жыл бұрын

Glad to hear that!

Salute To your Effort , Thanks a lot :)

Wonderfull . Explanation is awesome 👏👏👏

Tysm for crystal clear explanation.. God bless you

Bro, whatever comment section may say, I loved the video and understood it. Keep going👍

@CodeWrestling

3 жыл бұрын

Thank you so much 😀

what an amazing explanation..... thank you !!

@CodeWrestling

4 жыл бұрын

Glad you enjoyed it!

clear explanation sir we have understood easily but if we don't write it than again we forget after few days, so I have written from starting till end.Thanks for explaining clearly and in a simple manner.

Great explanation except the sound issues :) Background music is nice but should be a little lower :) Thank you for video man very helped for me :)

to implement backpropagation in neural network algorithm , is there any built in packages available as part of sci-kit learn ? Or do I have to write a script explicitly to implement back propagation ?

The video shows -with all the advantages as well as the limitations of working with a specific neural graph and particular numerical values- what is the BPP of errors in a feedforward network. But the basic idea applies to much more general cases. Several steps are involved. 1.- More general processing units. Any continuously differentiable function of inputs and weights will do; these inputs and weights can belong -beyond Euclidean spaces- to any Hilbert space. Derivatives are linear transformations and the derivative of a neural processing unit is the direct sum of its partial derivatives, with respect to the inputs and with respect to the weights; this is a linear transformation expressed as the sum of its restrictions to a pair of complementary subspaces. 2.- More general layers (any number of units). Single unit layers can create a bottleneck that renders the whole network useless. Putting together several units in a unique layer is equivalent to taking their product (as functions, in the sense of set theory). The layers are functions of the of inputs and of the weights of the totality of the units. The derivative of a layer is then the product of the derivatives of the units; this is a product of linear transformations. 3.- Networks with any number of layers. A network is the composition (as functions, and in the set theoretical sense) of its layers. By the chain rule the derivative of the network is the composition of the derivatives of the layers; this is a composition of linear transformations. 4.- Quadratic error of a function. --- Since this comment is becoming too long I will stop here. The point is that a very general viewpoint clarifies many aspects of BPP. If you are interested in the full story and have some familiarity with Hilbert spaces please Google for papers dealing with backpropagation in Hilbert spaces. For a glimpse into a deep learning algorithm which is orders of magnitude more efficient, controllable and faster that BPP search in this platform for a video about: deep learning without backpropagation. Daniel Crespin

Thank u so much ❤️

@CodeWrestling

4 жыл бұрын

Thanks for appreciating!! #CodeWrestling

Bohot khoob!!

Nice slides :)

Excellent video...nicely explained...

It was perfect. Also, to do not hear the sound, you may use a hands-free, it worked for me

excellent video, you have earned a suscriber

@CodeWrestling

3 жыл бұрын

Awesome, thank you!

Well explained the math behind it

Excellent.............Thanks

@CodeWrestling

Жыл бұрын

Thank you too!

Very clear explanation

@CodeWrestling

Жыл бұрын

Glad you liked it

really helpful and easy to understand thanks

@CodeWrestling

Жыл бұрын

You are welcome!

Thank u so much

Really good explanation

Bro inspite of using total error for w5 can we just use error of out 1.. Cz in hidden layer any how we have to use again error out1

I have a question qbout updating the weight. During using weight update formula, you have not used the multilayer perception rule in which we assume the +ve or -ve value of learning rate to increase or decrease weight. I have a doubt about this rule.

The music was too loud but your explanation was very helpful .

Best explanation ever!

@CodeWrestling

Жыл бұрын

Glad it was helpful!

thanks so much :)

Great explanation. It was really helpful. 😇

@CodeWrestling

3 жыл бұрын

Glad it was helpful!

hello sir you said you will implement decision tree algorithm in python. but i didn't find it in your playlist

If we give a one input and output, then weights will get adjusted as per those . But when another set of input and output are given , entirely new weights are formed. Then how does training happen in such algorithm?

superb ........

Should we also update the bias terms b1 and b2?

Very Nice, Vert Systematic

can we use ANN and fixed-effect Poisson regression model ? in two steps for better results ?

Bht sahi explanation

Thank You so much .. Finally i understood back propagation... Do you have video about Elman Recurrent Neural Network ?. If you have please send me your video link.

masterpiece

tq very much

good job

Very Nice

The song is to low, i can almost hear you. Make the song volume higher next time >3

@CodeWrestling

4 жыл бұрын

Sorry for the loud background music, we will make it better next time.

Awesome video thank you so much. Just became a subscriber. Back ground music is kinda high though.

@CodeWrestling

3 жыл бұрын

Thanks for the sub! Noted.

Happy teachers day

very nicely explained. It would be better without the background music.

Thanks

The first equation at 13:55 seems incorrect, can anyone confirm this?

Net h1 should be b1+(w1*x1)+(w3*X2) as per the network connections... Because it is w3 that is connected to h1 and not w2 Correct me if I'm wrong... Or else like my comment so that I would know

Nice video bro......but i had to watch it at .5 speed so as to follow along.

One of the best explanations of how backprop works. Better than those animated and fancy but boneless videos. Simple yet covers everything from scratch. Thanks and congrats to this guy!

@CodeWrestling

3 жыл бұрын

Wow, thanks!

Excellent

@CodeWrestling

4 жыл бұрын

thank you so much.

this vedio is simply super

@CodeWrestling

4 жыл бұрын

Thank you so much 😀

Fascinating BGM

HOW DID YOU GET THE LEARNING RATE CONSTANT? IS IT GIVEN IN QUESN?

Why did you take mean square error ? Neural network is basically a connection of many logistic regression output , So wouldn't it be the log loss function ?

hi, from where 0.1384 came in last?

very good explanation! but the music is a bit loud

@10:52 , I am unable to understand net_o1 = w_5 * out_h1 + w_6 * out_h2+ b_2 * 1. how does the derivative translate into 1 * out_h1 * w_5^(1 - 1) + 0 + 0 = out_h1 = 0.593269992, from where did w_5^(1-1) come from shouldnt it just be 1 * out_h1

@amitsahoo1989

4 жыл бұрын

exponent of x decreses by one in derivatives ..i think that is what he is trying to show ,..

background music was bit annoying.... couldnt hear or focus on what you were saying

Great introduciton to BNN , Nonetheless the background Music is a little bit loudy .

you didn't explain where the -1 came from (at 9:26 ) in the formula for dEtotal=2 * 1/2(targeto1-outo1)^2-1 * -1 + 0 . the derivative rules for power don't consider any `-1` values

@14:15 should there be w2*i2 instead of w3*i2 in the formula of net_h1?

@anonymous-random

4 жыл бұрын

you're right

why don't you update bias?

Thank your for creating the video. A small feedback will be that, the background music is too loud, I bearealy can understand what you are saying because of the volume of the background music. I

saw hell lotta videos and blogs including mattmazzur blog.... But you ripped like anything , bro...Amazing explanation, awesome job....literally you saved my time & made me to catch sleep early.....but how did you get 1/1+e-x as e^x/ a_e^x

@CodeWrestling

3 жыл бұрын

Thanks a ton. If you will differentiate it, you will get the and. Maybe quickly google it and you will get the answer.

thank you for uploading video with f in youtuber music , i couldn't focus shit but good content tho

can u send the ppt format of the concept

Why do you add music 🎶 in the background, we want the content not the music, it's too irritating....

Bro plz do other concepts of machine learning. I'm not getting wt my lecture is teaching ,I'm depending on your vedios .can u plz do make vedios regularly

Shouldn’t 1/2 = 1/n where n = the number of nodes for the givem layer? Really the sum of all errors for given layer is just the average. BTW great explanation.

I need Matlab coding of exactly this manual work bro😐