Neural Networks from Scratch - P.2 Coding a Layer

Expanding from a single neuron with 3 inputs to a layer of neurons with 4 inputs.
Neural Networks from Scratch book: nnfs.io
Playlist for this series: • Neural Networks from S...
Python 3 basics: pythonprogramming.net/introdu...
Intermediate Python (w/ OOP): pythonprogramming.net/introdu...
Mug link for fellow mug aficionados: amzn.to/2Vz9Hs0
Channel membership: / @sentdex
Discord: / discord
Support the content: pythonprogramming.net/support...
Twitter: / sentdex
Instagram: / sentdex
Facebook: / pythonprogramming.net
Twitch: / sentdex
#nnfs #python #neuralnetworks

[@mankaransingh2599]

Other People: Some Netflix Series Me: SENTDEX TUTORIAL SERIES !!

[@SanataniAryavrat]

hahaha... yeah, you're right!

[@milanpandey8431]

Nailed it!

[@StarkPlayz]

damn right!!

[@USAIsrUKEUVngrdBLRckOccupiedUA]

Very true! :))))))

[@oludhe7]

This would be funny if it wasn't true

[@robertschumann6977]

Again, I just want to thank you for this great series, especially those animations are amazing. Providing free education on such a high level is something what people nowadays more and more forget to value...

[@sentdex]

You're very welcome!

[@danielstatler954]

Imagine people 400 years ago. This level of knowledge would only be available to the lord's and anyone who knows it would price it so high

[@Joe_Zajac]

Still extremely helpful 3 years later!

[@meunomejaestavaemuso]

I tried coding a NN from scratch a few years ago, feed-foward is pretty easy, the tricky bit is back-propagation, looking forward to that.

[@zahhym]

I know its a nightmare

[@Foodius]

@@zahhym If you haven't watched it yet, you might want to look at 3blue1brown's series on neural networks. It's relatively math heavy but it really helped me to understand backpropagation more intuitively

[@zahhym]

@@Foodius I have. Its now just less of a nightmare :)

[@vaibhavkhobragade9773]

@@zahhym I can't disagree.

[@Jordan....]

Keep looking forward then.

[@ahmadayazamin3313]

"Neural Networks from Scratch" I expected a tutorial on building neural networks using x86 assembly...

[@sevret313]

It could be worse, it could be NN with Scratch.

[@mobob7989]

Red literal opposite level of abstraction haha

[@ahbarahad3203]

TFW Ben Eater posts a video

[@tornadospin9]

sentdex is so skilled at teaching that I bet he could make a course on that and it would be super easy to follow along

[@whoisabishag3433]

@@sevret313 ... kzbin.info/www/bejne/m3uUeZeKj6aWpJY

[@JoeElliotSA]

This is the course I have been waiting for. I have tried a few courses and learned so much BUT never fully understood what exactly I was doing 🤣. I always wanted to know what was behind the deep learning algorithms. Many thanks for this series of videos. 👍

[@sentdex]

Happy to make em!

[@ekremdincel1505]

@@sentdex Hi Sentdex! Please upload those videos fast :) I cant wait.

How did you put the python logo in your name lol

[@AdityaPratapSingh-ss8wg]

@ thats because he has join this sentdex channel...

[@SanataniAryavrat]

agree 100% ... he SentDex bro, looking forward to next video soon....

[@danielai7138]

The way you explain plus the animations from manim makes it so pleasant to watch. Thanks for this awesome content, this is what we all have been waiting for.

[@SanataniAryavrat]

Agree... animation made the whole video more interesting and engaging one...

[@godbennett]

Truly godly, using 313b's manim tool :)

[@godbennett]

I also wrote a free neural network book, based on 313b's neural network series: www.researchgate.net/publication/321162382_Artificial_Neural_Nets_For_Kids I ended up making a nice large backprop slice image in paint, so I didn't end up using manim :) imgur.com/NjT9EDt

[@jordivilahdl]

Which tool can I use to generate similar animations?

[@userre85]

Manim

[@adrianwuthrich982]

This is the first time I activated notifications for a channel

[@muhammadharis7581]

same here

[@parsafakhar]

i'm so exited about this series, god knows how many years i wanted something like this, thx sentdex i really appreciate it

[@sentdex]

Happy to do it!

[@soldour7597]

no one would believe if I say neural network could be explained in such way...but there you go putting it out! stay cool :)

[@johna30]

I'm loving these videos! Cutting away to the smooth flow animations/diagrams is very helpful in understanding what's going on. Thank you!

[@tranbao2799]

Looking at the animation for each line, I know how much dedication this guy has on this topic. Respect ✊

[@leonlysak4927]

Dude the ability you have for breaking down the complex into simple is nothing short of amazing.

[@charimuvilla8693]

Can't wait for the backprop part. Other youtubers confused me but I'm sure you will help me undrestand everything as always hehe

[@khmsalhuas1]

I am looking forward to finally really grasping the science behind deep learning. In the past, I just used libraries without understanding much. Great Series so far. Waiting for part 3.

[@sentdex]

Working on P3 as we speak!

[@dippy9119]

Just dropping a comment to show my appreciation for this video series. You have a real gift for teaching. This is by miles the clearest explanation I have found on how to build an artificial neural network. Am definitely interested in the book now.

[@italiandarthvader]

I have been searching for countless videos on KZbin to start with deep learning and this is the first one I found where it is explained in a way I can understand. I'm two videos in but I just want to say you made me fall in love with deep learning. Thank you!

[@BizzNitzz]

Love how easy to make this! After years of trying to learn this and taking a break then coming back I find it easier to learn thanks to you!

[@dumbtex6107]

your stuff is just so clear and concise i've done the whole keras tensorflow combo and theres just a huge gap in understanding behind this stuff and this dead basics stuff is just the perfect way to represent what's happening great stuff dood

[@PatrickHoodDaniel]

Your scaffolding approach to this topic is great. Thank you for making this series.

[@KingDuken]

Thank you making this super easy. This quite literally the best explanation I’ve ever seen. I have absolutely zero knowledge of how code a neural network before this point. Looking forward to getting into more in the series.

[@zakariaabdisalam1728]

sentdex, there are many tutorials online on neural networks, and I swear I had the same problem which was to hardcode the process and visualize the process from neuron to neuron and from layer to layer. Non of the NN tuts that I've watched went over this. However, you started to make an amazing tut with detail visualization and hardcoded NN process, basically you filled up all the gaps for learning. And you definitely deserve more than me joining the membership. Love you So much continue doing your amazing tuts. Looking forward to buying your book. ❤❤❤❤

[@danieloliveira9815]

It is just incredible the amount of effort that he puts in these videos, omg, thank you soooo much for making such a quality content available for free. You are really making the difference, man

[@sam_sam_sam]

I've tried other neural net and machine learning videos and pretty much all of them dive into a library and gloss over what the library is actually doing in the background. Thanks so much for explaining in this detail

[@RoboEvercool]

This is such an amazing thing that you and Daniel are doing. I especially love that you make the draft available to the community to make inputs and ultimately improve the overall final product! Such an awesome idea.

[@sentdex]

Glad you enjoy it!

[@vinamrakhoria1330]

omg i have been waiting for part 2 ever since part 1 came out thank you so much for this much needed playlist sentdex :)

[@sentdex]

You're welcome!

[@farhadnasiri3329]

Your teaching style is so understandable and straight forward... I hope you continue to do so

[@LunarMtnStudios]

You're very inspiring! I just bought the book and I'm excited to get into it. I'm a beginner python programmer but I've spent a lot of hours watching your tutorials. Maybe one day I"ll get to work with you!

[@insected83]

one of the best resources I've encountered for this subject. I have no computer science background and work in vfx. I am teaching myself python and machine learning to complement my market/equities trading and have been diving deep into this new path. This series (and channel) has really empowered my learning but more importantly my motivation and mindset. Instead of feeling overwhelmed I feel excited to keep learning. Thankyou sendtex, you are changing lives :)

[@subhrangshuadhikary6115]

Really appreciate your work. Please finish the series soon. Can't wait to learn all these soon.

[@fernandomaluenda4226]

Pretty incredible how simple you can make it. Thank you so much!

[@adenzu]

The animation is flawless really makes it easy to track and understand the process

[@anshmishra143]

This is a fantastic series. Great explanation and animations. Looking forward to release of further parts. Thanks a ton for such a great set of videos.

[@navis_6383]

This series is like premium quality.... It goes even further than paid ones.. You have my respect sir for making it free for us... I am just new to Neural network.. And I am so glad I found your video... A sub from me

[@thecathode]

This is brilliant, the animations are awesome. I wish I had this when I studied neural networks back in 1993; at the time I quit and chose another very interesting topic (digital filters) and never finished my backpropagation implementation in C++ which I found the other day in a diskette., which I think will be fun to revisit.

[@Random-my5rv]

Nothing but love for this guy! The tutorials we've all been waiting for!

[@alejandroalejandre3512]

Yo. Your tutorials are legit man. Very easy to understand and you explain complex processes clearly

[@santanumahata3956]

Such an awesome teacher who always thinks about the students getting the idea. Thanks a lot, sir.

[@spencerpalladino2604]

I am really loving this series. I feel like I am learning alot coding along with you.

[@lucasfernando4655]

The way you teach is so simple. Thank you so much!

[@jurkokubik6383]

This is awesome, so thank you for making this series! It’s super easy to understand. I’m really hoping that you continue make more and as much as you can. Honestly it would be super cool if you cool make these more often.

[@sentdex]

If they could be made any faster, I would :P They take a very long time to make.

[@jurkokubik6383]

sentdex well then I will just thank you for putting so much work into these and for being a good teacher.

[@shambhaviaggarwal9977]

I had just started learning neural networks and watched first episode of this series few hours ago and i wished god that i want to see more of this. And MIRACLE happened!!

[@alishaik6029]

What a great explanation man, really appreciate the edit and teaching. Thanks for teaching this much clearly.

[@karimloberg6876]

I am so eager to learn as fast as possible so I can understand everything that is being taught on this channel. Really nice content and personality.

[@astrolearnings2030]

I am used to guys not knowing what they are blabbering about but I can understand the concepts you are talking about quite clearly so thanks for that Mann pure gold content.

[@lerneninverschiedenenforme7513]

I want more commercials on youtube like this! After watching, I have a strong voice in my head that tells me to buy the book... and that was even before 13:53; Just because you mentioned "as in the book" in the beginning of the video. In short: I really like that you go so slow on these basics, that other people / tutorials fail to explain properly! Also, I really liked, that you did not use a loop! All credits for that!

[@gokulrejithkumar9612]

So far so good, I just hope you keep it this simple in the upcoming videos!

[@codetastic4749]

i have been trying to find a tutorial that actually teaches and doesnt just say DO THIS. this is really helpful. I actually understand this a bit better. thank you.

[@luxaslolete]

Amazing editing! You can tell theres a lot detail and attention put into the animations, cheers.

[@Tiashe0]

You explain this so fluently that its really easy to follow through. cant wait for the next one. Also i preordered the book.

[@sentdex]

Awesome! thank you for your support!

[@mattymallz4207]

Thanks for the best tutorial out there right now! Keep up the great work. The credits too, i loved it.

[@sentdex]

Thanks, will do!

[@emmanuelgrebshtein6435]

Are you planning to cover the new TensorFlow quantum library? It sounds really interesting. In any case, thanks for what you're doing!

[@sentdex]

Possibly, but right now the focus is here.

[@gabrielcorrea730]

@@sentdex Can you make a video just to show its power?

[@re.liable]

I really love those animations you do in the code.

[@Lleanlleawrg]

Thanks for doing this low-level series, it's very easy to follow and it's so basic, it's easy to learn, because everything is already familiar, allowing me to connect the ideas into an understandable concept.

[@sentdex]

Glad you enjoy it!

[@TrevorTheITGuy]

cant wait for the book!!! so many projects that ill be able to custom build the nn for

[@ulissemini5492]

the animations are on point! I was able to figure out the why there are N weights and 1 bias per neruon by slowing down the nerual network animation from last video, I'm so glad you addressed it for the people unable to figure it out, i'm sure you saved many other people a few hours of confusion :D

[@sentdex]

Glad I could help!

[@LuxurioMusic]

Honestly, I've not been able to get far with learning python, in fact I'd given up many months ago. However these two videos have helped incredibly, the clarity in the way you teach in amazing.

[@haktwld]

This is the most digestible beginning to networks. Thanks!

[@NeuralxAi]

This Series will be the best one eveeeer Created In Neural Network World !!! Love it man ❤😍

[@hadianasliwa]

Currently I'm reading "Deep Learning from Scratch: Building with Python from First Principles by Seth" but it seems somehow complex from the very beginning, but your video realy helps me go through the book. And it seems your book to be much easier to follow!

[@John-xi2im]

very awesome tutorial! explanation using both code and diagrams is the best part !

[@jaken82]

Amazing tutorial very easy to follow. Essential for anyone into ANNs.

[@AlanPope]

This is perfectly paced. I appreciate some may want you to go faster, or add loops at this stage. You're doing the right thing by starting properly from scratch! Nice work.

[@sentdex]

Thank you, glad you're finding this slow pace helpful! No one left behind!

[@terensedaman4608]

Honestly, amazing teaching skills.

[@bejgaonnischit9718]

I am totally looking forward to this series

[@kristiyanivanov7414]

Loved the patiently explaining!

[@walidr]

Hi Sentdex, thank you for starting and making this amazing course. It would be cool if you added a challenges of different difficulties at the end of each video. For example, for this one, you can challenge things like: 1) easy: write for 5 inputs to 3 outputs, 2) medium: rewrite using numpy (you will probably do this next video, but its still nice to think about before hand). This can help solidify the knowledge and also prepare us for the next video. Maybe the hardest challenge can be something to prepare for the next video. Keep up the amazing work!

[@clementsiow176]

I watched the 1 ep and now I am so excited for the whole series

[@junealexissantos4341]

This is the best tutorial ive watched so far. Thank you for this. This is very helpful to me who doesnt have any knowledge in coding and data science.

[@sentdex]

Glad it was helpful!

[@Mreoinpaul]

Brilliant explanatory animation of using the inputs, weights, and bias at about 2.15

[@tekleyoutube3259]

eagerly waiting all of your series tutorials. Because am sure that they will help me build my thesis on lung cancer classification, preprocessing and detection accurately

[@nishantsvnit]

That is really cool. This is a classical CNN problem to begin with and I hope you develop it and succeed with your goals.

[@tekleyoutube3259]

@@nishantsvnit sure am working on it. Though am facing challenges on segmentation and preprocessing all the datasets of luna16 and kaggle data science bowl 2017 stage 1. I will be glad if any one helps me tackle this challenge.

[@snharish5981]

Yeaaaa I've been waiting for this

[@ninjahunter2199]

i really like this visualization explaining the code of where the inputs and weights are coming from

[@craigrichards5472]

Great work mate, can't wait to get the book.

[@mohitjain4943]

No one can Teach with that ease and simplicity.....i really thank you man! this pace helps a lot Thanks Man!

[@sentdex]

Glad it helped!

[@He_Who_Could_be_Drunk]

Thanks for this great tutorial. I've been trying to get started with neural networks and attempting to understand them. I've been following along in C++ but this has been a really great reference for this. Thanks again!!

[@krishnakompalli2606]

Finally, a course worth watching. Everything is detailed properly as expected. The book is a must-buy but sadly I (many) can't afford😅. I hope there will be discount or give-aways.

[@kasiditauable]

This is the best tutorial for implementing neural networks using python. I should've found this before going for master in data science.

[@captaincode4380]

Great! I love this course! Its simple and I can understand it.

[@oludhe7]

This is the best thing about quarantine. These videos can't come out fast enough!

[@shamanbhattacharyya9285]

Love your videos. Your videos are far superior to what lecturers at universities can show

[@ginxiankok5045]

This is how I tried to code a NN about maybe 11 years ago, lol. Great stuffs.

[@karanmishra8192]

Your videos are always awesome and i have seen your all videos .Thanks for helping us by making videos on these topics..big fan of your work..

[@MJ46.91]

I think starting with the math concept behind this would've helped alot more than being willing to put this humongous effort, thank you very much, you've taught me alot through the years,

[@blackburn116]

I love you sentdex! I've learn so much with your videos. Thank you so much for keeping up your awesome work!

[@sentdex]

Happy to do it!

[@kubo-svk790]

Can't wait for book to arrive 😄 Keep up the good work. You are MVP for many university students and hobiests 😄

[@afterhours39]

Until now, the term "Neural Network" has been a mystery to me, despite doing a fair amount of research into it. By starting from scratch when teaching it, you have helped me understand the basic core concepts. "Thank you" just doesn't express my gratitude. THANK YOU!

[@sentdex]

Great to hear!

[@abdurrasheed2938]

What a wonderful explanation and presentation.. Thanks for sharing 🌹

[@nishitdua]

Been using keras for around a month now and tried building my own perceptron, Damn was that hard to do, all the complexity is hidden in the black box, am so glad i did it though. Now YT recommended me.Really cool for anyone who wants Learn how NN works under the hood. and i definitely recommend to everyone to not just watch this but to go ahead and try it own your and improve on the structure and move ahead of the tutorial.

[@mehdighorbani6008]

Explanation is AMAZING!!!! Thank you very much😍😍😍

[@federicopinto9353]

great content! The animations and the clear approach makes it easy to understand.

[@sentdex]

Glad you like them!

[@David-xy4nw]

you are the definition of a legend

[@amrhelal991]

i wounder why you don't have millions of subscribers , the best youtube channel so far............ keep up the good work greetings from Egypt

[@khizaridrees1758]

The person with no hate, your channel has everything from basic to advance. whenever someone asks for programming guide, I say one word: sentdex is your man

[@Narain346]

Daunting work sensei...keep going...we need it.🙏

[@spicytuna08]

awesome animation with those equations. so easy to follow. thank you so much.

[@sammflynn6751]

sir,you have one of THE most awesome channel in youtube. A speacial request ,please resume carla and self driving stuff

[@egorkostan2060]

Finally starting to understand something. Please Keep going. Thanks you for your effort.

[@8o8inSquares]

Couldn't be more clear. Well done! TO THE NEXT ONE!

[@jeanjohnson9737]

Nice content :) very much excited for this series.

[@tobias1204]

I love these animations! The style of visualisation reminds me of 3Blue1Brown.

[@sentdex]

That's because it's code we built on top of manim: github.com/3b1b/manim :D