Tagged on:                                                                 

100 thoughts on “But what is a Neural Network? | Deep learning, chapter 1

  • September 5, 2019 at 8:38 am
    Permalink

    number of spots in the hidden layer shouldn't be more ??

    Reply
  • September 6, 2019 at 9:34 am
    Permalink

    The son of Lisha li and sander would've IQ of 300.

    Reply
  • September 7, 2019 at 4:49 pm
    Permalink

    You're a god, Grant.

    Reply
  • September 7, 2019 at 5:08 pm
    Permalink

    bacd Answer

    Reply
  • September 7, 2019 at 5:58 pm
    Permalink

    That's certainly not how the brain and its neurons work but human is creating a way to convert things into numbers and numbers into another meaning number. That's good for machine to process data but I'm not sure if it can become a new intelligence or not. Let' s see.

    Reply
  • September 8, 2019 at 2:24 pm
    Permalink

    Hey, werent the beginning test sample numbers used to train an Neural network made of a sheet of glass ?

    Reply
  • September 8, 2019 at 7:16 pm
    Permalink

    Will you please suggest me a book to learn machine learning, please?

    Reply
  • September 9, 2019 at 4:25 pm
    Permalink

    Il campo elettromagnetico umano e' come quello del pianeta il cervello come universo

    Reply
  • September 10, 2019 at 12:45 pm
    Permalink

    Amazing

    Reply
  • September 10, 2019 at 1:46 pm
    Permalink

    i understood everything
    but i am having feeling that i just don't remember anything 😂
    is it normal?

    this is my first time watching a video from 3blue1brown.
    and i just become a fan. thanx! 😊

    Reply
  • September 10, 2019 at 1:51 pm
    Permalink

    Very nicely explained

    Reply
  • September 10, 2019 at 6:53 pm
    Permalink

    How could anyone downvote this? Bar none, the best neural net intro vid series ever.

    Reply
  • September 11, 2019 at 10:23 am
    Permalink

    🗽

    Reply
  • September 11, 2019 at 5:50 pm
    Permalink

    It is an awesome explanation. Thanks so much! One thing to correct, i believe the bias vector is till Bk instead of Bn.

    Reply
  • September 12, 2019 at 1:32 am
    Permalink

    … wenn jeder mit jedem vernetzt ist – und so in kürzester Zeit untereinander Infos ausgetauscht werden können.

    Reply
  • September 12, 2019 at 3:19 pm
    Permalink

    わっかりやすぅい

    Reply
  • September 12, 2019 at 6:56 pm
    Permalink

    If you have multiple layers with equal number of nodes, then several sets of weights can give identical results. With two hidden layers of equal number of nodes, one can permute the weights of the first layer, and give the inverse permutation on the next layer. Is that a useful property? I'm new to this except for an incomprehensible lecture I attended 30 years ago. This permutation symmetry may explain why you don't see the little pieces of numerals in the first layer you got. The solving mechanism is blind to the fact that permutations could give the same results. There will be however one particular permutation (or a few close ones) that will allow you so see the bits. But this will be a few in 16 factorial.

    Reply
  • September 13, 2019 at 10:47 am
    Permalink

    14:40 The biases here should range from b[0] to b[k], not from b[0] to b[n], I'm pretty sure.

    Since the result of the matrix vector multiplication will be a K by 1 matrix and not a N by 1 one, and moreover we have K neurons in the current layer, therefore K biases.

    Reply
  • September 14, 2019 at 5:03 pm
    Permalink

    중간에 한글자막끝남

    Reply
  • September 15, 2019 at 3:56 am
    Permalink

    Regards, awesome explanation.

    Reply
  • September 15, 2019 at 4:55 am
    Permalink

    this was some goddamn flame

    Reply
  • September 15, 2019 at 5:34 am
    Permalink

    神经网络跟神经系统有2毛钱关系

    Reply
  • September 15, 2019 at 11:39 am
    Permalink

    After watching the entire video just the thought of recognising objects in images blew my mind. Never appreciated image recognition more.

    Reply
  • September 16, 2019 at 7:29 pm
    Permalink

    No one in the world explains at this level!!!! 🤩

    Reply
  • September 17, 2019 at 4:33 am
    Permalink

    What you say is how we view and recognize a given digit, is not how we do it. I could explain it here, but there are plenty of good books on the subject.

    Reply
  • September 17, 2019 at 4:50 am
    Permalink

    a 2D Neuronal Network, lol. 3 D Chips are still out of the question, because of heat problems.

    Reply
  • September 18, 2019 at 2:37 am
    Permalink

    I feel like someone stole this and made "lets make my home town into a brain"

    Reply
  • September 18, 2019 at 10:03 am
    Permalink

    https://www.youtube.com/playlist?list=PLvf_ZGHboMr9DZ750qpW3G1dCKpXkDpvv

    Reply
  • September 18, 2019 at 12:37 pm
    Permalink

    Good afternoon Mr.Patreon sir and thank you sir for your good teaching on neural network and so many good things.

    Reply
  • September 19, 2019 at 9:55 am
    Permalink

    14:52 mistake: bk instead of bn

    Reply
  • September 20, 2019 at 9:26 am
    Permalink

    I'm not really impressed, if this is what "Machine Learning" is – this is algorithms and divide and conquer style allows computers to 'Read' hand writing, there is a name for it – OCR which is technology that has been around for over 10 years now (this is used in Scanning Software). So are you SURE this is machine Learning?

    Reply
  • September 20, 2019 at 1:10 pm
    Permalink

    I have created a web application that supports your learning process in the field of neural networks.

    https://blueneurons.ch/nn

    You can play around with the settings (architecture, activation functions, training settings) and observe how the settings affect the predictions. All datasets have preconfigured values that can be adopted. It is also possible to create your own datasets.

    Instructions and explanations to the implemented elements:

    https://blueneurons.ch/pdfs/userguide.pdf

    Reply
  • September 20, 2019 at 4:11 pm
    Permalink

    YOU ARE AWSOME! I saw 10 videos to find what they are. I simply understood by watching the first 7 minutes of yours.

    Reply
  • September 21, 2019 at 3:10 pm
    Permalink

    이제 기계한테 사람이 인식하는 과정 비슷하게 하도록 했으니까 다음 인식과정을 추측해서 우리가 배우면 되겠다.

    Reply
  • September 22, 2019 at 6:01 am
    Permalink

    What is deep learning? A branch of pseudoscience.

    Reply
  • September 22, 2019 at 8:08 pm
    Permalink

    We're GODS.

    Reply
  • September 23, 2019 at 3:20 am
    Permalink

    Anybody else sometimes feel the need to clap at the end of the video?

    Reply
  • September 23, 2019 at 3:40 am
    Permalink

    Thank you for all your awesome videos! I have learned so much about mathematics via your visual representations. Keep up the awesome work. I have started a YouTube of my own now. I Just uploaded my very first video! 😊👍

    Reply
  • September 23, 2019 at 2:43 pm
    Permalink

    Didn't understand the weights part. What are they for?

    Reply
  • September 23, 2019 at 6:07 pm
    Permalink

    Wow I love you!

    Reply
  • September 24, 2019 at 3:14 pm
    Permalink

    Best video about ANN i have ever seen… so clear and precise! Good job 👍

    Reply
  • September 24, 2019 at 11:39 pm
    Permalink

    Thank you! This helps a lot and it also opens up sooo many questions!! Guess that means it's on to the next video 🙂

    Reply
  • September 26, 2019 at 9:14 am
    Permalink

    @0:56 the video says "and outputs a single number between zero and ten". It should say "between zero and nine". -Annoying commenter.

    Reply
  • September 26, 2019 at 10:48 am
    Permalink

    Best tutor on Youtube

    Reply
  • September 27, 2019 at 6:26 am
    Permalink

    Captcha has left the chat

    Reply
  • September 27, 2019 at 1:55 pm
    Permalink

    Sir can you please include videos on other machine learning algorithms,image processing and NLP..it will be very useful for people like us who are enthusiastic to learn these in a conceptual way.

    Reply
  • September 28, 2019 at 12:26 pm
    Permalink

    How else would we read doctors writing.

    Reply
  • September 29, 2019 at 2:34 am
    Permalink

    Best frickin' videos on complex things. You're an amazing teacher.

    Reply
  • September 29, 2019 at 4:58 pm
    Permalink

    So it's like "Match the following " kinda thing right ?

    Reply
  • September 29, 2019 at 11:49 pm
    Permalink

    Thanks for this awesome content!!

    Reply
  • September 30, 2019 at 7:12 pm
    Permalink

    Ben:ne?
    Beynime:bak gördün mü
    Artık daha fazla değer vermeliyim nöronlara

    Reply
  • October 1, 2019 at 1:31 pm
    Permalink

    Fantastic….developed intuition

    Reply
  • October 1, 2019 at 1:36 pm
    Permalink

    My doctor's prescription would be the best test for the Neural Network

    Reply
  • October 1, 2019 at 2:51 pm
    Permalink

    бля я же английский не понимаю. зачем старался понять вижуал. но всё же мои нейроны смогли что то понять. странно

    Reply
  • October 1, 2019 at 8:36 pm
    Permalink

    Sorry but this method seems archaic to solve AI

    Reply
  • October 2, 2019 at 1:51 am
    Permalink

    Thank you so much for Japanese subtitle.

    Reply
  • October 2, 2019 at 7:10 pm
    Permalink

    artificially intelligence gives same pattern as we practice jyotish in India

    Reply
  • October 2, 2019 at 7:12 pm
    Permalink

    we use factors of planet's and astral body so we used that in predictions of randomness about anything in the universe

    Reply
  • October 2, 2019 at 7:18 pm
    Permalink

    if artificially intelligence used all the equation of jyotish from India then we easily produced that structures which gives actual intelligence to machines like it gives an average person to think about which is not recognise by anyone but it's going to be happened in past or future

    Reply
  • October 2, 2019 at 7:21 pm
    Permalink

    use coral draw and photo shop for recognise the picture

    Reply
  • October 2, 2019 at 7:23 pm
    Permalink

    use the property of trigonometry about the language symbols so we got the basic idea about any symbols

    Reply
  • October 2, 2019 at 8:13 pm
    Permalink

    Чувак, у нас в русском есть З.И это не цифра. Так что не пиЗди лишнего.

    Reply
  • October 3, 2019 at 2:45 am
    Permalink

    wtf

    Reply
  • October 3, 2019 at 6:37 pm
    Permalink

    lindo, incrível, bonito e gostosão

    Reply
  • October 4, 2019 at 7:12 am
    Permalink

    The thumbnail is moved when i scroll down fast

    Reply
  • October 5, 2019 at 6:41 am
    Permalink

    não entendi vo me mata

    Reply
  • October 5, 2019 at 9:12 pm
    Permalink

    Cool!

    Reply
  • October 6, 2019 at 10:27 am
    Permalink

    Well that escalated quickly

    Reply
  • October 6, 2019 at 3:40 pm
    Permalink

    Love the video. Also loved the idea to depict weights as red / green pixels. My eyes are sore though. The problem is that I am colorblind, green tries to elude me. Please try to colorblid-proof your videos 🙂

    Reply
  • October 7, 2019 at 12:37 am
    Permalink

    why 16 weights?

    Reply
  • October 7, 2019 at 9:14 am
    Permalink

    Brain: "I'm a stupid human beeing.."
    Neuron: "Hold my 0 to 1!"

    Reply
  • October 7, 2019 at 9:18 am
    Permalink

    Please watch Robert Sapolsky's "Behave" lecture it gets into the two methods of how the visual cortex processes images. There's a cognitive method but it's slow, and a fast method which is really lighting fast and in accurate as heck. not something that you want when someone is running at you during a riot carrying cell phone.

    Reply
  • October 7, 2019 at 9:26 am
    Permalink

    Neurons most common bullying phrase:

    "You just dont have any value to me!"

    Reply
  • October 7, 2019 at 9:30 am
    Permalink

    Would it be possible to make a software OOP body with multiple AI embed in the body class with every parts of the brain compartments for objects with a neural network function in each object?could an AI or neural net be used to automate the population of the nodes in the parts of the brain compartments objects?
    weights and biases sounds like opthamalogy theory in linguistics.

    Reply
  • October 7, 2019 at 8:45 pm
    Permalink

    Thanks this helped much more than my class there was just a whole bunch of equations on the slides -_-

    Reply
  • October 7, 2019 at 11:36 pm
    Permalink

    Still people thinks they are created by an accident… Just think of your mind and learning capabilities….

    Reply
  • October 8, 2019 at 3:47 am
    Permalink

    Hey , I just have one small question
    The neuron which we were discussing above as a function is basically a perceptron rt??

    Reply
  • October 8, 2019 at 8:52 pm
    Permalink

    Is this the Reuben Report?

    Reply
  • October 9, 2019 at 4:33 am
    Permalink

    Leisha is hot. If you had babies with her, they would solve the universe. So get on it!

    Reply
  • October 9, 2019 at 5:58 am
    Permalink

    Whoo it is a very clear explanation!

    Reply
  • October 9, 2019 at 12:43 pm
    Permalink

    Please make more videos about machine learning and data science. I wish I could have a teacher like this at the school.

    Reply
  • October 10, 2019 at 8:58 am
    Permalink

    Hi, how do you recognize the edges of the font?

    Do you recognize automatically, or, have they been divided before?

    Thank you

    ric

    Reply
  • October 10, 2019 at 11:56 am
    Permalink

    This is where need for quantum computers kicks in

    Reply
  • October 10, 2019 at 1:47 pm
    Permalink

    15:13 LOL !
    BTW great work boss ! idk how much work would have gone in making a quality video like this !

    Reply
  • October 10, 2019 at 1:49 pm
    Permalink

    great way of convincing people to subscribe ! LOL

    Reply
  • October 11, 2019 at 8:38 pm
    Permalink

    My circuits circle in my mind and get hot and never learn. Promise, they will never learn.

    Reply
  • October 11, 2019 at 8:43 pm
    Permalink

    📂TheFuck

    └📁What

    Reply
  • October 12, 2019 at 6:40 am
    Permalink

    I can't wait for neural networking to be able to recognize my doctor's prescription.

    Reply
  • October 12, 2019 at 4:09 pm
    Permalink

    I watched this video a while ago it's very good. Having watched a lot of other videos on your channel I think you could do very well explaining neural networks as differentiable tensor operations more generally

    Reply
  • October 12, 2019 at 4:27 pm
    Permalink

    Dude, this explanation is facking amazing, thanks for it)))

    Reply
  • October 12, 2019 at 6:23 pm
    Permalink

    Hmm.

    Reply
  • October 13, 2019 at 9:23 am
    Permalink

    how is it possible that I can lie in my bed on a Sunday and am presented with mind-boggling cutting edge knowledge told by an incredibly soothing voice in a world class manner on a 2K screen of a pocket supercomputer basically for free

    Reply
  • October 13, 2019 at 2:25 pm
    Permalink

    the way this woman talks hurts my ears

    Reply
  • October 13, 2019 at 5:15 pm
    Permalink

    Bardzo ciekawe .

    Reply
  • October 14, 2019 at 3:11 am
    Permalink

    thx

    Reply
  • October 15, 2019 at 12:24 am
    Permalink

    If bias differ between each layer, does that increase dimensionality? Are biases trained as well?

    Reply
  • October 15, 2019 at 10:33 am
    Permalink

    I'm a 14 years old non english fluent and i understood everything. Is it normal? Is there anyone here like me?

    Reply
  • October 15, 2019 at 3:28 pm
    Permalink

    I am enlightened. I finally saw the light last night after watching this video thousand times. Thank you sincerely.

    Reply
  • October 16, 2019 at 10:25 pm
    Permalink

    You are the best! And Squishification lol 😂

    Reply
  • October 17, 2019 at 5:10 pm
    Permalink

    so neural network is actually a function and choosing right values for its 13,000 parameters ?

    and for each set of problems these parameter values will be changed?

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *