How Deep Neural Networks Work

[@flavialan4544]

This should be recommended as the 1st video to watch when it comes to learn neural networks

[@DR-bq4ph]

Yes

[@ckpioo]

yes I agree but for simplicity sake he should have done a 0 to 1, 0 being black 1 being white and 0.5 being grey, because almost everyone follows that pattern, and for new learners its a bit harder for them to switch from thinking about -1 to 1 to 0 to 1

[@SimonZimmermann82]

totally agree. i understood the principle right away and now we have our own nn in operation

[@heyasmusic7553]

I watched your videos 3 years ago. It'salmost nostalgic. You may not see this. But you're one of the reasons I kept moving through with Machine Learning

[@BrandonRohrer]

I legit cried a little bit. Thank you for this.

[@danklabunde]

I've been struggling to wrap my head around this topic for a few days, now. You went through everything very slowly and thoroughly and I'm now ready to dive into more complex lessons on this. Thank you so much, Brandon!

[@BrandonRohrer]

I'm very happy to hear it :)

[@mikewen8216]

I've watched many videos and read many blogs and articles, you are literally the best explainer at making these intuitive to understand

[@behrampatel3563]

I agree.Penny dropped for me today with this Video. Thank you so much Brandon

[@a.yashwanth]

3blue1brown

[@biokult7828]

"Connections are weighted, MEANING".... Holy fuck.....after viewing numerous videos from youtube, online courses and google talks.... (often with comments below saying "thanks for the clear explanation")....This is the FIRST person i have EVER seen that has actually explained what the purpose of weights are....

[@Tremor244]

I feel the same, even though I still can't completely understand how weighting works :/

[@garretthart4883]

Tremor244 I am by no means an expert but weighting is what makes the network "learn" to be correct. By changing the weights it changes the output of each neuron and eventually the output of the network. If you tune the weights enough you will eventually get an output that is what it is supposed to be. i hope this helps

[@LuxSolari]

I don't work with neural networks but with other types of machine learning. But weighting is more or less the same in all these fields of mathematics. You want a system that, provided with an input (an image, for instance), achieves its classification as the output. For instance you have a scenery (input) and you want to know if it's from vacations at the mountains or at the beach (a classification, ie. the output). So you pass the image trough a set of filters: (1) does the image have umbrellas? (2) does it have clouds? (3) is there a lot of blue? (4) is there a lot of brown?, etc. If the image passes a specific combination of filters, there is a greater probability that the image is of a specific type (for instance, if the image (1) have umbrellas, (3) is blueish and isn't (4) brownish, it's more likely to be from the BEACH). But how much more likely? That's when the WEIGHTING comes into play. Through machine learning we want to calculate some coefficients (weights) that state a sort of likelihood of an image to pass a filter, given its type (for instance, if it has umbrellas there's a probability of 0.9 out of 1 (90%) that it is from the beach and not from a mountain, but if there's a lot of blue maybe only 0.6 of those images are from the beach, and so the WEIGHT IS LIGHTER. That means that, if the image passes a filter of COLOR BLUE it is likely to be from a BEACH, but if it passes a filter of UMBRELLAS it is EVEN MORE LIKELY). Weights, then, are a parameter of RELEVANCE of each of the selected filters to achieve the correct classification. So we make the machine learn from LOTS (thousands, perhaps) of images that we KNOW are from the beach or the mountain. One image from the beach has umbrellas, so the classification through the filters was correct and then the WEIGHT for the umbrellas is increased. But if there is an image of the mountains with umbrellas and the program says it's from the beach, the weight goes down for the umbrellas. When we did this with a lot of images, the weights are FINE TUNED to classify correctly most of the time (if the filters are any good... if we chose wrong filters from the beginning, then there's a chance the dictionary won't get any better even fed with lots of images. That could also happen if the training images are biased: ie. if they don't represent the real set of images that we want to classify). I hope this works better for you!

[@anselmoufc]

If you have had a course on linear regression, you will recognize weights are equivalent to parameters. They are just "free variables" you adjust in order to match inputs with outputs. In one-dimensional linear regression, the parameters are the slope and offset of a line, you adjust them so that the distance between the line and your points (your training examples) is the least. Neural networks use the same idea as statistical regression. The main difference is that neural networks use a lot of weights (parameters), and for this reason you have to care about overfitting. This in general does not happen in linear regression, since the models are way more parsimonious (use only a few parameters). The use of a lot of weights is also the reason why neural networks are good general approximators, the large amount of weights give them high flexibility. They are like bazookas, while statistical regression is more like a small gun. The point is that most of the times you need only a small gun. However, people like to apply neural networks to problems where linear regression would do a good job since NN are "sexier".

[@madsbjerg8186]

+Esteban Lucas Solari I want to let you know that I love you for everything you just wrote.

[@mukulbarai82]

I've watched many videos on KZbin but non of the videos explained the concepts as intuitively as you did. Thought I have to watch it again as I've failed to grasp some concepts, I am sure that it will be clear as I watch more.

[@BrandonRohrer]

Thanks Mukul!

[@claireanderson5903]

Brilliant! I was involved 50 years ago in a very early AI project and was exposed to simple neural nets back then. Of course, having no need for neural nets, I forgot most of what I ever knew about them during the interval. And, wow, has the field expanded since then. You have given a very clear and accessible explanation of deep networks and their workings. Will happily subscribe and hope to find further edification on Reinforcement Learning from you. THANK YOU.

[@klaudialustig3259]

I already knew how neural networks work, but next time someone asks me, I'll consider showing him or her this video! Your explanation is visualized really nicely.

[@jones1351]

Fantastic description of what these networks do. I've gone thru a few of these explainers and all they demonstrated was the person knew their subject, they just couldn't teach it. They talk in jargon, that quickly loses those unfamiliar. IOW they're not teaching, they're having a 'conversation' with those who are already versed and have background in the field. Einstein is to have said, 'If you can't explain it simply, then you don't understand it yourself' Thanks, again. I walk away feeling like I actually learned something. You Can Teach.

[@BrandonRohrer]

Hey thanks! I really appreciate this. It's the highest compliment.

[@MatthewKleinsmith]

Great video. Here are my notes: 7:54: The edges going into the bottom right node should be white instead of black. This small error repeats throughout the video. 10:47: You fixed the color error. 11:15: Man, this video feels good. 21:41: Man, this video feels really good. An extension for the interested: Sometimes we calculate the error of a network not by comparing its output to labels immediately, but by first putting its output through a function, and comparing that new output to something we consider to be the truth. That function could be another neural network. For example, in real-time style transfer (Johnson et al.), the network we train takes an image and transforms it into another image; we then take that generated image and analyze it with another neural network, comparing the new output with something we consider to be the truth. The point of the second neural network is to assess the error in the generated image in a deeper way than just calculating errors pixel by pixel with respect to an image we consider to be the truth. The authors of the real-time style transfer paper call this higher-level error "perceptual loss", as opposed to "per-pixel loss". I know this was outside the scope of this video, but it was helpful to me to write it, and I hope it will help someone who reads it.

[@humanity3.090]

Good to know that I'm not the only one who caught the logical mistakes. 9:14 Bottom second squash should be vertically inverted, if I'm not mistaken.

[@ganondorfchampin]

I had the idea of doing perceptual loss before I even knew the term for it, seems like it would work better for warp transforms and the like versus level transforms.

[@hozelda]

Alternatively, the edges are correct but the corresponding picture should be flipped. Regardless, the final step (output perceptron at the bottom indicating horizontal) works with either the white white edges or the black black edges scenario.

[@oz459]

thanks :)

[@sali-math-arts2769]

YES - thanks , I saw that tiny error too 🙂

[@andrewschroeder4167]

I hate how many people try to explain complicated concepts that require math without using math. Because you used clear mathematical notation, you made this much easier to understand. Thank you so much.

[@cloudywithachanceofparticl2321]

A physics guy coming into coding, this video completely clarified the topic. Your treatment of this topic is perfect!

Don't worry people I asked this guy if he was a physicist

[@Mau365PP]

@ thanks bro

[@FlashKenTutorials]

Clean, concise, informative, astonishingly helpful, you have my deepest gratitude.

[@BrandonRohrer]

You are most welcome

[@intros1854]

Finally! You are the only one on the internet who explained this properly!

[@rickiehatchell8637]

Clean, concise, informative, astonishingly helpful, you have my deepest gratitude. I've never seen anyone explain backprop as well as you just did, great job!

[@fghj-zh6cv]

This simple lecture truly makes all viewers fully understand the logic behind neural networks. I strongly recommend this video clip to my colleagues participated in data driven industry. Thanks.

[@InsaneAssassin24]

As a chemist who just recently took Physical Chemistry, back propagation makes SOOO much more sense to me when you put it into a calculus description, rather than a qualitative one as I've been seeing elsewhere. So THANK YOU!

[@jabrilsdev]

this is probably the best breakdown ive came across, very dense, you've left no spaces in between your explanations! Thanks for the great lesson! Onward to a calculus class!

[@abhimanyusingh4281]

I have been trying develop a DNN for a week. I have seen almost a 100 videos, forums, blogs. Of all those this is the only one with calculus that made complete sense to me. You sir are the real MVP

[@NewMediaServicesDe]

30 years ago, I studied computer science. we were into pattern-recognition and stuff, and I was always interested in learning machines, but couldn't get the underlying principle. now, I got it. that was simply brilliant. thanks a lot.

[@bowbert23]

I always had trouble intuitvely understanding how a derivate works and how practically its calculation is reflected in simple terms. Little did I know starting this video, that I'll finally understand it. Thank you! I'm relieved and feel less stupid now.

[@BrandonRohrer]

I'm really happy to hear that Bowbert. Thank you for the note.

[@thehoxgenre]

i was amazed by the way you talk, and explain very slowly as well you remain slow until the end and you dont rush things. bravo

[@alignedbyprinciple]

I have seen many many videos regarding NN but this is by far the best; Brandon understands the relationship between the NN and the backbone of the NN, which is the underlining math. He clearly presented them in a very intuitive way. Hats off for you sir. Keep up the good job.

[@sirnate9065]

Who else paused the video at 15:10, went and did a semester of calculus, then came back and finished watching?

[@muhammedsalih4846]

Nobody

[@danielschwegler5220]

:)

[@danielschwegler5220]

Muhammed Sahli's mother

[@safesploit]

SirNate I still remember most of my calculus and have notes from prior study 😜

[@SreenikethanI]

lol

[@DeltaTrader]

Possibly one of the best explanations about NN out there... Congratulations!

[@Gunth0r]

My kind of teacher! Subscribed! Nice voice, nice face, nice tempo, nice amount of information, nice visuals. You'd almost start to believe this video was produced with the concepts you've talked about. And my mind was just blown. I realized that we could make a lot more types of virtual neurons and in that way outclass our own brains (at even a fraction of the informational capacity) with a multitude of task-specific sub-brains forming a higher brain that may or may not develop personality.

[@ViralKiller]

That was incredible...watched 7 videos so far and every day my brain understands a bit more...I recently learning Houdini VEX code which is 3D graphics programming, and that took 1 year of watching a whole bunch of stuff and not getting it...until I did...so I know I will grasp this soon....Im sticking to these simple examples for now, until I can code it from scratch in Python

[@radioactium]

Wow, this is a very simple explanation, and it helped me understand the concept of neural networks. Thank you.

[@SunyangFu]

The best and easily understandable neural net video I have seen

[@marioeraso3674]

Awesome description of what neural networks are!

[@tomryan9827]

Great video. A single clear, concrete example is more useful than 100 articles full of abstract equations and brushed-over details. Speaking as someone who's read 100 articles full of abstract equations and brushed-over details.

[@Toonfish_]

I've never seen anyone explain backprop as well as you just did, great job!

[@ViralKiller]

I never understood backprop properly until this video...this was the light bulb

[@user-kr6dk7bq6b]

It's the first time I get to understand how neural networks work. Thank you.

[@yashsharma6112]

Very very rare way to explain a neural network in such a great depth. Loved the way you explained it ❤

[@AnkitSharma-ir8ud]

Really great explanation Brandon. Also, I greatly appreciate that you share your slides as well and that too in raw (PPT) format. Great work.

[@dexmoe]

Very detailed and clear explanation. Thank you for sharing! :)

[@OtRatsaphong]

Thank you Brandon for taking the time to explain the logic behind neural networks. You have given me enough information to take the next steps towards building one of my own... and thank you KZbin algo for bringing this video to my attention.

[@bestoonhussien2851]

I'm in love with the way you explain things! So professional yet simple and easy to follow. Keep it up!

[@Sascha8a]

This is a really good video! For me as a complete beginner this really help me understand the basics of neural networks, thanks!

[@AviPars]

Artem Kovera lovely book , just downloaded. for the lazy people : amzn.to/2ntC9Zm

[@DanielMoleGuacamole]

Holy thank you!! ive watched like 50+ ich tutorials on neural networks but all of em explained things poorly or too fast. But you went through everything slowly and actually explained all the info clearly!!

[@BrandonRohrer]

Thank you so much! I'm happy to hear how helpful it was, and it means a lot that you would send me a note saying so.

[@coolcasper3]

This is the most intuitive explanation of neutral nets that I've seen, keep up the great content!

[@Jojooo64]

Best video explaining neural networks i found so far. Thank you a lot!

[@RichaChauhandreams]

@Brandon Rohrer To each neuron 1. First a number is assigned 2. Then a weight is assigned 3. Then it is squashed using sigmoid and then the values are summed up! Right ? My question is whether each weighted neuron is squashed or the sum of weighted neurons is squashed , why is sqaushing done?

[@BrandonRohrer]

Good question Richa, and it will take a bit longer to answer. There's a deeper dive into this material at e2eml.school/312

[@lucazarts25]

OMG it's even harder then I expected! Thank you very much for the thorough and thoughtful explanation!

[@lucazarts25]

it goes without saying that I became a subscriber as well ;)

[@PierreThierryKPH]

Very slowly and clearly gets to the point, nice and accessible video on the subject.

[@salmamohsen8208]

Easiest most elaborate explanation I have found on that matter

[@abhijeetbhowmik2264]

The best Back Propagation explanation on you tube. Thank you sir.

[@abubakar205]

one of the best teacher you cleared all my doubts for neural networks thanks sir let me click an ad for you

[@yassinelamarti4157]

the best explanation for Neural Networks ever !

[@mdellertson]

Yours was a very easy explanation of deep neural networks. Each step in the process was broken down into bite-sized chunks, making it very clear what's going on inside a deep neural network. Thanks so much!

[@shahidmahmood7252]

Superb!! The best explanation of DL that I have come across after completing the Andrew NG's Stanford ML course. I am a follower now.

[@antwonmccadney5994]

Holy shit! Now I... I actually get it! Thank you! Clean, concise, informative, astonishingly helpful, you have my deepest gratitude.

[@zw0503]

It took me trhee goes in excel to replicate the idea, but ... after eliminating my mistakes i now have a working model . my next step will be to generalise this to make any network 2X2 3X3 or 5X6 . thanks your explanation is/was very clear

[@BrandonRohrer]

Awesome Robert. I love to hear it.

[@Mr_AciD]

At 7:48, the bottom right receptive field should be Black black white white, not White white black black :) Congratulations on the explanation!

[@yhr4052]

Yes, there is a mistake.

[@BrandonRohrer]

It is true! Good catch both of you.

[@vipinsingh-dj2ty]

literally THE best explanation i found on the internet.

[@junepark1003]

This is one of the best explanations I’ve come across. Thank you! And subscribed :)

[@dbiswas]

You explanation is so far the best. I am sure you are the best teacher so far. Thanks for uploading such a informative video.

[@MarcusHammarberg]

Math, baby! It’s cool

[@cveja69]

I almost never post comments, but this one deserve it :D Truly great :D

[@technoultimategaming2999]

I understand every topic. From computer enginerring to calculus and AI basics, but put them all together and working to make a digital functioning brain is magic

[@GAV32]

Thank you so much! I have been trying to create a neural network of my own for simple tasks, and I haven't been able to learn how until now. Thank you!

[@TanNguyen-vm2fc]

Gavin Haynes please teach me how to start. im a beginner of this field. thank you

[@GAV32]

I haven't made one of my own yet, sorry for the confusion. The difference is that now, I understand how one works, so I can start structuring my information.

[@AngryCanadian3]

This is one of the best explanation of neural network I have seen

[@antoinedorman]

This is gold if your looking to learn neural networks!! Well done

[@kd4pba]

Damn, This is perfect. I had kept backing away from all this until now. I finally get it. You really helped me Brandon, thank you.

[@srinivasabugada2726]

You explained How Neural Networks in very simple and easy to understand manner. thanks for sharing!

[@ziaurrehman2180]

After wasting too much time finally I found the right place , excellent explanation 👏👏👏💯

[@surtmcgert5087]

so I ahve a quick questions, In the network you said there was negative weights(black) and positive weights(wight) so when your making the network do you pre-define which weights should be positive and negative or do they work themselves out when you train the network?

[@BrandonRohrer]

They work themselves out during training. If you'd like to go deep on the topic, there's a course on it with detailed tutorials and code: e2eml.school/321

[@surtmcgert5087]

@@BrandonRohrer thank you very much, this has cleared alot of stuff up

[@cheaterman49]

I totally want to implement this now. Can't be that hard - won't be the best approach, just trying to see how close it can "naturally" get to the ideal solution you displayed and how many iterations of training it takes!

[@BrandonRohrer]

Give it a go! Although I would love to get a sticker for my laptop that says "Can't be that hard." I'll put it right above the one that says "What could possibly go wrong?"

[@fulanomengano8895]

I've been trying to implement the full network as seen (@ 25:30) in python but hit a roadblock. Have you done it?

[@rahimdehkharghani]

This is the best explanation I have ever seen for neural networks. thanks very much!

[@ahuttee]

You should check out 3blue1brown's series on neural networks, its absolutely beautiful

[@opeller]

Thank you so much, really helped me understand several things that were hard to understand during class.

[@BrandonRohrer]

I'm glad to hear it!

[@tobiaskarl4939]

Excellently explained ! Automatic subtitle feature enabled would have been nice.

[@BrandonRohrer]

Thanks Tobias! Subtitles for English (and a dozen other languages!) are enabled. I hope they work for you now.

[@tobiaskarl4939]

@@BrandonRohrer yes, thx a lot.

[@MadScientist267]

Finally a useful approach... A "real" (complete) network. The query mechanism makes perfect sense now. Finally. Lol Now if I could just get there with the backward stuff. It makes enough sense to know it will eventually make perfect sense but still struggling a little there. You've got one of the best explanations for beginners out there tho. I'm the type that has to see what the hell is actually going on... Don't need it later when a functioning whatever is running, but early on broken down like this is perfect because then the scale is just scale.

[@BrandonRohrer]

Thanks! If you haven't had a look yet, this may help with the backprop: kzbin.info/www/bejne/bHOwqJypib95qJY

[@Ivan_1791]

Best explanation I have seen so far man. Congratulations!

[@WilsonMar1]

I've seen a lot of videos and this is the most clear explanation. Exceptional graphics too.

[@Prof_C]

Thanks for making this creative commons. I am going to adapt and use for my class, with attribution. I have a number of project management resources that are creative commons if you need.

[@BrandonRohrer]

Thanks Scott. I hope this one serves you and your class well.

[@BrianKeeganMusic]

This makes sense and all, but how do neural networks make sense in terms of prediction modeling for say a regression problem? Not image classification. Like housing prices for example?

[@alfakannan]

You are a gifted teacher. Even I could understand.

[@shivamkeshri487]

wow awesome i never find a video like this with the simple example and clarity of neural network and its a though topic to explain but you make it easy... thanks

[@nirbhaythacker6662]

The function shown at 4:39 and 20:33 are both referred to as being the same function, but the graph on 4:39 is actually 2*(o(a)-0.5) , here o(a) is the sigmoid function.

[@PotatoMan1491]

Really helped me contextualise the matrix procedure in practical sense. Thank you!

[@kademmohammed6836]

by far the best video about ANN i've watched, thank you so much, really clear

[@Thejosiphas]

I like how much effort you put into making these ideas accessible

[@Eddie-rf4tp]

Very interesting.. I learned this from the top down with tensorflow. What I am wanting to learn, is about different node types. Keep at it, it's good to see all the youtube instructors these days.

[@BrandonRohrer]

Thanks Eddie. Will do.

[@MavVRX]

Would it be better to train a very deep neural network in small chunks as opposed to just adding a lot of layers and letting back prop do the job?

[@liamdev4855]

The *Best* Video for beginners!

[@BrandonRohrer]

Thanks!

[@ZeroRelevance]

Great video, just one question. At roughly 7:56, where you show the shaded squares, in the bottom right it is a bit confusing. The far bottom right one is taking negative input values, which, going off what you said, means it should have black on the top and white on the bottom.

[@Vermilicious]

Nice intro. Fairly easy to grasp the essence.

[@nilaier1430]

Watching this video on my 4 pixel screen phone. Really informative.

[@lordgoessl]

Thank you for your awesome videos! I learned a lot about neural networks lately and your videos helped me a lot. Only because of this video i was able to finish a neural network in Java based on your video and example! I just tried it with numerical gradient estimation and it trains to a cost under 1e-18 your with 8 training sets (perfect solid, perfect diagional, perfect vertical, perfect horizontal and everything inversed) in about 5 seconds! I'm still adjusting the steps it makes when changing the weights but it works! Now I'm trying to implement the mathematical gradient descent but this will also be a lot easier thanks to your videos! (I know, English is difficult)

[@BrandonRohrer]

Sebastian, that is fantastic. I'm pleased that the explanations helped. If you feel like sharing a copy of your code, I'm sure a lot of viewers would be interested. I've gotten a lot of requests for a working example. If you decide to throw it into a GitHub repository, let me know and I'll point to it from my blog.

[@lordgoessl]

Sure, I just cleaned everything a little bit because I'm only a student so I have never published code. Would it be ok if i use the example in your video? I would mention your name, video, KZbin link and blog link?

[@BrandonRohrer]

But of course! I'm glad it can be of use. Kudos on sharing your code.

[@lordgoessl]

Great thank you! I uploaded it to GitHub! github.com/sebig3000/MachineLearning The training process is still a little crappy because it just tries to adjust every weight one at the time and the usses the less costy adjusting. So it's basically a finite difference approach. But I'm working on the direct gradient calculation!

[@jacolansac]

The internet needed a video like this one. Thanks a lot!

[@davidguaita]

You're the man at explaining these things. Thank you so much.

[@slayemin]

This explanation of back propagation was exactly what I needed. This is very clear and I now have higher confidence in my ability to create my own ANN from scratch.

[@mehranmemnai]

Same here. My vision is clear

[@brendawilliams8062]

I just enjoy numbers. Anything to do with them is a fantastic thing.

[@thangvu-uj7qi]

I watched this video in very early days of my PhD. Now, it appears in my recommendation list when i have just graduated. Such a nice memory ❤

[@BrandonRohrer]

That is such a good story. Thanks thắng.

[@Beudd]

This video is crazy good. Truly, this is amazingly well explained from the beginning till the end. Wow, thanks a lot for such an excellent presentation.

[@SyedMehdiX]

That was flat out the best video explaining neural networks. Thank you!

[@daksheshgupta7045]

I've a question, how do we determine the number of hidden layers and nodes for a layer in a neural network ?

[@BrandonRohrer]

That's a big question. Here's a course I wrote to help answer it: e2eml.school/314

[@daksheshgupta7045]

@@BrandonRohrer Thanks, I'll check it out !

[@jhwblender]

I've been wanting to make a neural network for some time. I knew it required calculus, and I have taken calculus. The mathematics others have explained seemed way over my head. But you have made it very clear. Thank you!

[@brendawilliams8062]

I must know calculas now. Yes!

[@fajarrukmo]

Only by seeing thumbnail i can understand the concept. But still don't know how to implement in software. Do we need create function for every neuron and send result to other function until we can calculate the weight in the end ??

[@BrandonRohrer]

It's true there are a lot of fine details that still need to be filled in. To get the full story, including a Python implementation, check out End to End Machine Learning Course 312. e2eml.school/312

[@garretthart4883]

This video is hands down the best intro to neural networks I've ever seen! fantastic job. And thank you for putting links to learn more and not just leaving us hanging. I think it also just solidifies that you know what your talking about. I look forward to more content from you. Good work!

[@snehotoshbanerjee1938]

Your all videos on NN are excellent!

[@jonathanfowler2932]

Congratulations! You've earned a subscriber.

[@MilanNedic94]

Another one!

[@akostrikov]

Another one

[@MegaLovinglight]

another one :)

[@nightshadefns]

Another one! This is old :D

[@tonyxing9925]

one more

[@Squash101]

Is there some type of playlist for learning about neural networks?

[@BrandonRohrer]

Give this a try: kzbin.info/www/bejne/f33Wcmekrqxqf5I