5.1: Doodle Classifier: Introduction - Intelligence and Learning

[@jacksonabernathy1683]

Great vid man. Just discovered you recently, already my favorite channel on KZbin

[@geoffwagner4935]

Woot! nice shot, just finished toy neural . well for what it is. obviously intend on spending much more time pulling it's pieces apart again later , really like most of this . who knew this was literally next on the playset

[@aasimbaig01]

Omg cant wait for this series ... ♡♡♡

[@TheCodingTrain]

(They are now available unlisted if you click through the description!)

[@Alty950]

Yes! Can't wait to see more!

[@XiSparks]

5:56 What a circle!! :O

[@QuocNguyen-mh7ow]

Congrat you for 1M subcribers. Your videos are the best tutorial to learn Machine learning.

[@grainfrizz]

I'm sorry you had to re-upload, Dan. But I will still rewatch it no matter what.

[@Kombo-Chapfika]

Glad to be back!

[@Alessandro-nq3tm]

Amazing, thank you Dan for your hard work

[@TheCodingTrain]

Thank you!

[@SubjektDelta]

2:50 hahaha that "cello" above your head looks nice ;D

[@Bopas2]

I hope to follow along and complete my own version of this project! Very excited! Are you planning to go over how actually extract google's data, I looked around and couldn't figure that out. Thanks!

[@TheCodingTrain]

Click through to the next video (it's in the description below)!

[@Chris-Brown-]

I think I found a oversight in your original neural network When the hidden layer uses the original weights to find the output_error used during the feedforward instead of the error adjusted weights it resolves quicker

[@dappermink]

So, I am trying to make a digit classifier since one week in Python (and yes I watched all of the 3blue1brown videos about it). I did all the structure, have the mnist data to train it, and everything is ready except the back propagation part (the most important one lol). I was so happy to see you are about to code a very similar neural network. But sadly I understood you will use a librairy which will hide us the most important part which I needed. Hopefully though, I'm not fully done watching your playlist about Neural Networks so maybe your videos on the topic will help me :) Anyway, keep filling KZbin with Educational masterpieces please

[@TheCodingTrain]

In that playlist I step through building the entire neural network library!

[@dappermink]

The Coding Train Perfect then, I will need it ^^

[@НатальяСоколова-я1г]

Could you please help me find that video about building the entire neural network library?

[@dappermink]

Позитроныч Of course! There you go: kzbin.info/aero/PLRqwX-V7Uu6Y7MdSCaIfsxc561QI0U0Tb This is a playlist, not a single video, also you may want to take a look at the Session 3 if you have troubles at understanding some stuff he uses

[@vincentwhizin9282]

This series looks great, I’m super excited! Do you also have a series on the absolute basics for AI/Machine Learning in Python?

[@TheCodingTrain]

I don't unfortunately! But I hope to keep making AI/ML videos.

[@technicalilm8999]

github.com/pengoox/Doodle_classifier in python

[@nagesh007]

Awesome

[@kamilbolka]

Great! I need to find some time to do fun things like this :D

[@souravkar3211]

thanks dan, this is really gonna help me in developing some new projects

[@TheCodingTrain]

so glad to hear!

[@MrBHackerS]

What's the prior knowledge for this series ? Is there a link for previous videos series ?

[@TheCodingTrain]

Here are a collection of playlist that might help? kzbin.infoplaylists?shelf_id=16&view=50&sort=dd

[@lflee]

Dan, who are your favorite philosophers?

[@LorenzoLeonardini]

I am trying to make a doodle classifier, but with only a class: cat. How can I train a NN to say "ok this is a cat at 70%" without having other classes to compare with?

[@NunoLopes99]

13:00 has to become a meme!

[@goonfx]

can you make a tutorial for object detection & localization with TensorFlow.js pleaseeeee? :D

[@azulamazigh2789]

yes yes yes ... yeeeeeeeeeeeeeeeees

[@patrickpoggi5448]

Hey, may anyone help me coding Java on Atom on a Mac, please? I've just downloaded the JDK so far...

[@KanalMcLP]

Why do all your videos come out after i did the stuff by my self?

[@badhusha2693]

What kind of Neural Network is this, RNN or CNN , or any other thing??

[@TheCodingTrain]

This is just a plain vanilla neural network, I would like to do it with a CNN soon!

[@shgidi]

Is there a working demo?

[@TheCodingTrain]

See: github.com/CodingTrain/Toy-Neural-Network-JS/tree/master/examples/doodle_classification

[@jadestrider2660]

didn't think the note about "missing data sets" and "algorithmic violence" was necessary in this context.

[@georgecoghlan4663]

Just as I finish my course work on machine learning for image recognition. Typical.

[@joshinils]

Why the reupload?

[@TheCodingTrain]

There was an offensive word in the cello drawings I didn't notice. Apologies!

[@rcookie5128]

@Burger: if he did the whole job of removing & reuploading the video, it would be dumb to share it in the comments..

[@Naej7]

2:47

[@rcookie5128]

well like naej hinted, it's likely something that roughly shaped like a cello doodle..

[@Naej7]

ΜrCookie Not quite, someone started to draw, but got pissed off seeing a bad result, scraped the drawing and wrote an insult next to it

[@mix_tutorial_gaming2985]

System.out.println(Photogram.parseMoment(2:35[2][3]))

[@martlesmartlesmartles]

reupload...?

[@TheCodingTrain]

Yes, apologies, there was an unfortunate issue with the first upload I had to correct.

[@sainimohit23]

I wish these were in python

[@Ninjaznexx]

ಠ_ಠ mfw "algorithmic violence" at 10:15

[@jadestrider2660]

Ninjaznexx agreed; wtf.

[@TheMcallist1]

13:00 hahaha

[@isaacbaptista6207]

make a n original coding chanlage

[@monkeyrobotsinc.9875]

is it ethical? what kind of warning is that? youre giving me weird ideas. is that intended? can you gives us some examples? also, the missing data girl. is that for real? is she like an AI data miner or something? that looks too stupid to be real.

[@hendrik3553]

"draw animal migration" ..wut