Neural Networks Pt. 2: Backpropagation Main Ideas

[@statquest]

The full Neural Networks playlist, from the basics to deep learning, is here: kzbin.info/www/bejne/eaKyl5xqZrGZetk Support StatQuest by buying my book The StatQuest Illustrated Guide to Machine Learning or a Study Guide or Merch!!! statquest.org/statquest-store/

[@motherisape]

Bamm

[@gbchrs]

[@seanleith5312]

Quit the singing, please

[@statquest]

@@seanleith5312 Noted

[@dcscla]

Man, your promotions are not shameless! Actually, what you do is a gift for us, for the price that you charge and for the level of the content, we are being gifted and not a buying something. You are far better than a lot of paid (and expensive) courses. Just check out your video comments to see how people few happy when they discover your videos!! Great work as always. Thank you so much!!!👏🏻👏🏻👏🏻👏🏻

[@statquest]

Thank you very much! :)

[@Luxcium]

He is using the concept of reverse psychology by presenting great stuff at a good price and as you mentioned theses promotions are not shameless… They are shameful, as you hinted he should indeed be ashamed of giving us such a good and advantageous offer… 😅😅😅😅

[@jonforce360]

You released this video just in time for my AI exam! Thank you. Sometimes I think professors use really complex notation just to feel smarter than students, it doesn't help learning. I love your content.

[@statquest]

Thank you very much!

[@sarazahoor9133]

I want to copy-paste this comment! :D

[@puppergump4117]

Ain't that right. They must be mad that they don't understand the actually smart people so they don't want to be understood either.

[@zhongtianjackwang5346]

lol, that is exactly what I want to say

[@ElNick09]

I have been a student my entire life and have taught college level courses myself, and I must say you are one of the finest lecturers I have ever seen. This statquest is a gem. Your work is so succinct and clear its as much art as it is instruction. Thank you for this incredible resource!

[@statquest]

Thank you very much! :)

[@juliocerono_stone5365]

I am already 65, and your videos have helped me understand the basics behind NN. Thank you so much!!!!

[@statquest]

Bam! :)

[@iskrega]

I just want you to know your channel has been instrumental in helping me towards my Data Science degree, I'm currently in my last semester. I'll be forever grateful for your channel and the time you take to make these videos. Thank you so much.

[@statquest]

Thank you and good luck with your final semester! BAM! :)

[@hamzasaaran3011]

I am studying for a Master's degree in bioinformatics now, and as someone who knows little about statistics, I really can't thank you enough for your videos and the effort that you have put into them.

[@statquest]

Thank you!

[@akeslx]

I finished business school 25 years ago where I studied statistics and math. So happy to see that neural networks are fundamentally just a (much) more advanced regression analysis.

[@statquest]

BAM!!! Thank you for supporting StatQuest! Yes, neural networks are a lot like regression, but now we can fit non-linear shapes to the data, and we don't have to know in advance what that shape should be. Given enough activation functions and hidden layers, the neural network can figure it out on its own.

[@filipwojcik4133]

I've started my data science journey back in 2010. At that time, I struggled a lot to learn backpropagation for NN - I wanted to fully understand what's going on. Years passed, I worked in the data science industry giants, got my Ph.D. degree in the field meanwhile, and started teaching my own students. I can say one thing - StatQuest's backpropagation explanation is the best I've ever seen. I redirect all my students to this series of videos, as I believe there is no better, cleaner and no bull***t explanation available. Most of the videos and tutorials out there make it even more complicated by introducing **FANCY NOTATION** with deltas, nablas, etc. or other extra symbols. This one is simple, pure and gets the job done. Statquest - this is an excellent work, I wish we had more handbooks and papers written as clearly as your videos + books. All the best wishes!

[@statquest]

Thank you very much! I really appreciate it. I'm going to have a new book with all of these details about neural networks coming out in the next few months.

[@filipwojcik4133]

​@@statquest I'm really glad to hear that! Fingers crossed! It will be no 1 on my reading list, and recommended lectures for students as well :) Keep up the pace, that's awesome! TRIPLE BAM!!!!!

[@adirozeri7162]

You are amazing. THank you for taking this long road of showing us all the calculations. It really deepens my understanding and intuition!

[@statquest]

Thank you!

[@evangeliamm]

You have no idea how much I appreciate your work. Your explanations are so fun and simple, I'm just so grateful!

[@statquest]

Thank you very much! :)

[@yasameenmohammed4366]

My Machine Learning exam is tomorrow and re-watching your videos to review concepts is helping me so much! Thank you!!!

[@statquest]

Good luck! BAM! :)

[@Amir-gc8re]

Finally a proper, detailed, step by step explanation. This guy is absolutely AMAZING ! Thank you so much for all the hard work in putting these videos together for us.

[@statquest]

Thank you very much! :)

[@jays9591]

May I say .... You are such a good teacher that it is most enjoyable to watch your videos. I am proficient in statistics (via university econometrics 101) ... and I did not realise all those fancy terms in machine learning are actually concepts that are common items in the stats that I learned in the 1970s, e.g., biases and weights, label, activation functions etc. Anyway, I can see that a lot of viewers appreciate your work and teaching. I have also 'updated' myself. Thank you.

[@statquest]

Thank you very much!

[@babarali4313]

Its the teacher who makes the Subject easy or difficult and the way you explained Neural Network, I am speechless

[@statquest]

Thanks!

[@raunak5344]

I just iterated on a gradient descent and found that this is the best possible way to teach this topic and no other lecture in the entire existence is better than this one

[@statquest]

bam!

[@katwoods8514]

Love this! You've explained it far better than anywhere else I've seen, and you made it entertaining at the same time! Thank you so much for making this.

[@statquest]

Awesome, thank you!

[@advaithsahasranamam6170]

This is excellent stuff! As a visual learner, your channel is a BLESSING. Thank you so much for your fantastic work on breaking down concepts into small, bite-sized pieces. It's much less intimidating, and you deserve so much more appreciation . You also gained my subscription to your channel! Keep doing a great job, and thank you SO MUCH for having my back!

[@statquest]

Thank you very much!!! :)

[@mot7]

You are the best. I wish every ML learner find you first. I am going to do my part and tweet about you. Thanks for making these videos! Wish you more success.

[@statquest]

Wow! Thank you very much! I really appreciate the support. BAM! :)

[@ML-jx5zo]

Now Iam reading backpropagation, I worried about this vedio didn't came for long time , And finally I got a treasure.

[@statquest]

bam! :)

[@tagoreji2143]

Teaching such complicated topics in a simple, Easily Understandable way.👏👏👏.Thank you, Professor

[@statquest]

Thanks!

[@hyonnj9563]

Honestly you do a much better job at teaching using a pre recorded video than my instructors using both written and live materials that I'm paying for.

[@statquest]

I'm glad my videos are helpful! :)

[@edrobinson8248]

simply brilliant. Learning is indeed a quest. A quest for someone who understands and can present understandably. Thanks.

[@statquest]

Thanks!

[@TheClearwall]

Who else is using these videos to put together a semester project? So far, I've put Regression Trees, K-fold CV, complexity pruning, and now Neural networks as my final model construction. Josh is worth a double bam every time.

[@statquest]

BAM! Good luck with your project.

[@subusrable]

this video is a gem. I had to watch it a few times and like in gradient descent, I went closer to the target level of knowledge with each step :)

[@statquest]

BAM! :)

[@wong4359]

I found your explanation is far more easier to understand than the edx online course I am taking, BAM !!!

[@statquest]

bam!

[@ksrajavel]

Finally. The wait is overBAM!!!

[@statquest]

TRIPEL BAM!!!

[@mariolira9279]

F I F T H B A M!

[@syco_Rax]

SUPER BAM!!!

[@BlochSphere]

The level of detailing in this video is just 🤯 I hope i can try to make my Quantum Computing videos this clear!

[@statquest]

Good luck!

[@mohammadrahman1126]

Amazing explanation! I've spent years trying to learn this and it always went too quickly into the gory mathematical details. Aha moment for me was when green squiggle equal blue plus orange squiggles lol Thank you for this Josh!!!

[@statquest]

Glad it was helpful!

[@amandak1396]

Kind of like how Feyman reduced gory math in physics to actual squiggle, double bam!

[@katwoods8514]

omg yay! I just discovered that you've made a million videos on ML. I'm going to go binge all of them now :D

[@statquest]

Hope you enjoy!

[@motherisape]

Bamm

[@lukasaudir8]

I am really glad that people like you exist!! Thank you so much for those incredible lessons

[@statquest]

Glad you like them!

[@codeman2]

I searched neural net and again your video popped that too just 4 month old, love to get your helpful videos right before my semester

[@statquest]

:)

[@jblacktube]

I didn't even get through the jingle before I gave a thumbs up. Thanks for the chuckle, can't wait to watch the rest of this!

[@statquest]

BAM! :)

[@O5MO]

I never understood backpropagation. I knew some things from other tutorials, but as for beigginer, it was very hard to understand. This video (and probably series) is the best i could find. Thank you.

[@statquest]

Glad it was helpful!

[@jennystephens3215]

Josh, this is amazing. You really make things so easy to visualise which is crazy considering the hidden networks are meant to be so hard that they are referred to as black box! Thanks for all your videos. I have used heaps over the last twelve months. Thank you again.

[@statquest]

Hooray!!! I'm so glad that you like my videos. :)

[@ABCEE1000]

have no idea how can i thank you as you deserve .. thank you so much

[@statquest]

Thanks! :)

[@dinara8571]

JUST WOW! Thank you so much, Josh! I cannot express the feeling I had when EVERYTHING made sense!!! TRIPLE BAM! Never thought I would be extremely excited to pause the video and try to solve everything by hand before I look at the next steps

[@statquest]

BAM! :)

[@NadaaTaiyab]

oh that's a good idea!

[@AhmadAbuNassar]

Thank you very much for this comprehensive yet simple explanation

[@statquest]

Glad it was helpful!

[@mashmesh]

Omg, protect this man at all costs, this was pure gold!!! Also, thank you, sir, for talking so slowly because if my brain squiggles need to work faster they will burn up x)

[@statquest]

Glad you enjoyed it!

[@sarazahoor9133]

For the first time ever in history, I have understood the concept behind Neural Networks! BAM!!!! :D Thanks Josh, so grateful :)

[@statquest]

BAM! :)

[@manalisingh1128]

Wow Josh way to go!!!! You have the concepts so clear in your own head that it seems a piece of cake for us 🍰♥️ Love from India! 🇮🇳

[@statquest]

Thanks so much!!

[@madghostek3026]

9:00 at this moment I realised I'm watching the best math content on earth, because you never see simple stuff like this being given attention to. Luckily I already know how summation symbol works, but I didn't know it in the past, and nobody cared to explain. But it's just not about the summation symbol, imagine the other 1000 small things somebody might not understand, and doesn't realise they don't understand, because it's been skimmed over

[@statquest]

Thank you so much! I really appreciate it! :)

[@perhaps467]

Thank you so much for this series! I haven’t been able to find any other videos that really break down the mechanics of neural networks like this.

[@statquest]

Thanks!

[@user-re1bi2bc8b]

Incredible. Sometimes I need a refresher on these topics. There’s much to remember as a data scientist. I’m so glad I found your channel!

[@statquest]

Bam!

[@maliknauman3566]

How amazing is the way you convey complex concepts.

[@statquest]

Thank you!

[@edphi]

It should be made a crime for anyone to see other videos on backpropagation before they reach Statquest. The world is confused by teachers who tell the big story before the basic. Learn the basic and the picture fall into place like the chain rule 😊

[@statquest]

bam! :)

[@yiliu5403]

Best Neural Networks Lectures! Just ordered the book from Amazon to support!

[@statquest]

Wow! Than you very much! :)

[@iliasaarab7922]

Best explanation that I've seen so far on backpropagation!

[@statquest]

Thank you! :)

[@knt2112]

Hello sir, Thanks for such an simple explanation, never understood back propagation in such a depth at this ease. 🎉

[@statquest]

Thank you!

[@CHERKE_JEMA5575]

You rescued me from the unknown!! Much Love from Ethiopia

[@statquest]

Bam! :)

[@nojoodothmanal-ghamdi1026]

I . JUST . LOVE . YOUR . CHANNEL !! you literly explain things very clearly and step by step! I just cannot thank you enough really

[@statquest]

Wow, thank you!

[@aryabartarout5697]

You have cleared my doubt on back propagation, gradient descent and chain rule. Triple Bam !

[@statquest]

:)

[@David5005ful]

The type of in depth video I’ve always wanted!

[@statquest]

Thank you!

[@rodrigovm]

best teacher ever. Even better than Andrew Ng.

[@statquest]

Thank you!

[@ucanhnguyen4751]

Thank you for this video. I have been waiting for this all the time. Finally, it appeared just 1 day before my exam. You are a life saver!!

[@statquest]

Good luck with your exam! :)

[@terrepus9856]

The time couln't be more perfect ... 3 hours before my machine learning exam !! Thank you!!!!

[@statquest]

Good luck with your exam! I hope it goes well.

[@viethoalam9958]

give respect to my math teacher, but this is so much easier to understand.

[@statquest]

bam! :)

[@chrislee4531]

I learn more from four of your videos than 200 pages of textbook gibberish

[@statquest]

Thanks!

[@mjcampbell1183]

Wow! This is an incredible video. Thank you SO MUCH for making this for us. This is one of the best videos I've seen to explain this concept. The hard work you have put into this is something that I am incredibly appreciative of. Thanks, man.

[@statquest]

Wow, thank you!

[@superk9059]

Thank you very much for your video~ Your videos make me feel that studying English make so much sense, otherwise I can't enjoy such beautiful thing~ 👍👍👍❤❤❤

[@statquest]

WOW! Thank you very much!!! And thank you for your support!!! :)

[@josephif]

Lecture was awesome,more affective and easy to understand Thanks

[@statquest]

Thank you! :)

[@evie389]

I was reading an article based on Backpropagation and I did not understand a single word. I had to watch all your videos starting from Chain Rule, Gradient Descent, NNs...I re-read the article and understood everything!!! But now I can't get the beep--boop and small/double/triple/ bam out of my head lol.

[@statquest]

BAM! I'm glad my videos were helpful! :)

[@aminmoghaddam7624]

I wish our lecturers watched these videos before trying to make their own teaching slides! (With acknowledgement of course!)

[@statquest]

bam!

[@deepanjan1234]

This is really awesome. I thank you for your effort in developing this highly enriched content. BAM !!!

[@statquest]

Thank you!

[@willw4096]

Thanks for the great video!! My notes: 7:23 8:11 8:48 10:00 10:22❗,11:13 - 11:48, 11:56 12:08 13:30❗,

[@statquest]

BAM! :)

[@d_polymorpha]

Hello, thank you for the video! This series has been really helpful to learn about deep learning. I have a couple of questions. 1. When using gradient descent and backpropagation, do we always use SSR to measure how good a fit the parameter we are estimating is? Or are there other ways? 2. The second question is when using chain rule for calculating derivatives. The first part is d SSR/ d Predicted. In that first part @ 11:25 are you using chain rule again within that first part? And when deriving the inside Observed - Predicted @ 11:34 where do you get 0 and 1 from?

[@statquest]

1. The "loss function" we use for gradient descent depends on the problem we are trying to solve. In this case, we can use the SSR. However, another commonly used "loss function" is called Cross Entropy. You can learn more about cross entropy here: kzbin.info/www/bejne/bHLVhKypatZ7d7c and kzbin.info/www/bejne/rnOomWlsi56akNE 2. You can learn how the chain rule works (and understand the 0 and 1) here: kzbin.info/www/bejne/rZ2Unqyup9mEfrM

[@royazullay7556]

That Josh guy is just awsome !! Definitely will support !!

[@statquest]

Thank you!

[@danielkim4151]

Wow! Thank you so much for this video!!!! It helped me so much for my AI assignment.

[@statquest]

Glad it helped!

[@VishalKhopkar1296]

you taught this better than professors at CMU, not kidding

[@statquest]

Thank you! :)

[@Luxcium]

Wow 😮 I didn't knew I had to watch the *Gradient Descent Step-by-Step!!!* before I can watch the video related to *Neural Networks part 2* that I must watch before I can watch the *The StatQuest Introduction To PyTorch...* before I can watch the *Introduction to coding neural networks with PyTorch and Lightning* 🌩️ (it’s something related to the cloud I understand) I am genuinely so happy to learn about that stuff with you Josh❤ I will go watch the other videos first and then I will back propagate to this video...

[@statquest]

Getting warmer...

[@mrglootie101]

I've been waiting for this all the time checking the notification haha

[@statquest]

Hooray! The wait is over.

[@alexissanchezbro]

Your getting better and better. Thank you

[@alexissanchezbro]

BAAAAAAAMMM

[@statquest]

:)

[@utkugulgec5508]

These videos should be protected at all costs

[@statquest]

:)

[@epistemophilicmetalhead9454]

Back propagation (aka finding w's and b's) start with b_final=0. you'll notice that error = (actual - predicted)^2 is really high. so you find the gradient descent of squared error wrt b_final and find out the value of b_final for which the squared error is minimum. that is your optimal b_final. gradient descent: derivative of sum of squared errors wrt b_final = derivative of sum of squared errors wrt predicted value y * derivative of y wrt b_final. d(y observed - y predicted)^2/d(y predicted) = -2*(y observed - y predicted) d(y predicted)/d(b_final) = d(sum of all those previous curves obtained through each node of the layer + b_final)/d(b_final) = 0+0+0....+0+1=1 take the predicted curve ke x values and find the derivative/slope. step size = slope*learning rate. new b_final = old b_final - step size. keep repeating until slope touches 0. this is how gradient descent works and you've found your optimal b_final.

[@statquest]

double bam

[@amirhossientakeh5540]

perfect you explain complicated things very underatandable it's amazing

[@statquest]

Thank you very much! :)

[@aviknash]

Excellent job Josh!!! Just loved it!!! Thanks a ton for your fun-filled tutorials :)

[@statquest]

Glad you like them!

[@DanielRamBeats]

This is finally all making sense to me thank you

[@statquest]

Thanks!

[@alhaanali2502]

You got the best way to teach thank you❤

[@statquest]

Thanks!

[@ge13r]

Saludos desde San Cristóbal, Venezuela!!!

[@statquest]

:)

[@cthutu]

Excellent content, excellent delivery - just bought your book!

[@statquest]

Thank you so much for supporting StatQuest! BAM! :)

[@jieunboy]

insane teaching quality, thanks !

[@statquest]

Glad you think so!

[@zombieeplays3146]

I come to this channel for the intros tbh!

[@statquest]

bam! :)

[@marpin6162]

Thank you. Now everything is more clear.

[@statquest]

BAM! :)

[@JamesWasTakenOhWell]

Thank you for the amazing effort you put into this video and BAM!!! as always!

[@statquest]

Thanks!

[@preetikharb8283]

This video made my day, thank you so much, Josh!!

[@statquest]

Thanks!

[@constantthomas3830]

Thank you from France

[@statquest]

Merci! :)

[@harishbattula2672]

great explanation sir. Tripple BAM........kudos to your presentation.

[@statquest]

Thank you! :)

[@igorg4129]

Thanks Josh! you simply the best

[@statquest]

Thank you very much. I can't wait to get the other videos out soon.

[@Morais115]

I'm buying the shirt! Kudos to you sir.

[@statquest]

Awesome! Thank you!

[@gilao]

Another great one. Thanks!

[@statquest]

Thanks again!

[@mike___-fi5kp]

You always are the best.

[@statquest]

Thanks!

[@igorg4129]

Josh, finished watching. Thank you again 1 If I as a researcher know +/- which range of inputs I am going to insert, and which range of outputs I expect to get in the end, will I want to adjust somehow from the very beginning the weights range, maybe weights distribution, same thing about biases and same about activation functions, or today we let the algorithm to do this job? 2 most interesting question: Lets say that while finding the prediction curve we kind of discover some "hidden truth". I think our curve might never be exact also because we do not know all of the independent variables which in nature affect our dependent variable. Say we know one, but there is another one which we do not know about. If so, will it be right to say that when neural network with one input splits the input by different weights into two neurons of a hidden layer (from which the final output is calculated), it is like simulating somehow presence of another "secret independent variable" even without knowing what it is? Thanks

[@statquest]

I'll be honest, I'm not sure how to answer question #1. I don't know. I do know that some of the methods used for initializing the weights with random values increase the variation allowed in the values based on how many layers are in the neural network - so that might do the trick. As for the second question: Adding the second node in the hidden layer allows the squiggle to go up *and* go down. If I just had one node, I would only be able to go up *or* down. So, in some sense, that is sort of like adding a secret independent variable.

[@igorg4129]

@@statquest Also thought this way. Thank you again and again you do here a titanic job Josh. If not you I wasn't here to ask new questios. :)!

[@ZachariahRosenberg]

@@igorg4129 It's tempting to want to initialize weights to a target range in the hopes of speeding up convergence, however this actually might be counter productive. The weights of individual nodes do not have to conform to the same distribution as your output. When you use an appropriate (adaptive) optimizer, it should be able to tune the weights pretty quickly, considering that the first few passes will likely have larger gradients.

[@pranjalpatil9659]

Perfect explanation!

[@statquest]

Thank you!

[@harshpahade2703]

Hey, thanks for this playlist, it was a great find. Just pointing out a typo at 1:44 where the Note says 'spit' instead of 'split'.

[@miriza2]

BAM! Thanks Josh! You’re the best! Got myself a pink T-shirt 😍😍😍

[@statquest]

Hooray! And thank you for supporting StatQuest!!!

[@richarda1630]

Where were you 5 years ago???!?!?! :D Awesome work man! Keep it up :)

[@statquest]

Thanks! I have 4 more neural network videos coming out in the next month.

[@richarda1630]

@@statquest awesome! can't wait :D

[@SPLICY]

The understated BAM at 4:40 cracked me up 😂

[@statquest]

SPLICY in the house!!! BAM! :)

[@davidlu1003]

I think I understand the neural network clearly now.😁😁😁

[@statquest]

bam! :)

[@ssanand3]

I wish someday you make a video in person so that we can see the saint behind the voice 😀

[@statquest]

:)

[@zhenyuhe1537]

You can imagine how complicated professors explain it when even stat quest has to use equations

[@statquest]

:)