Machine Learning Zero to Hero (Google I/O'19)
Вставка
- Опубліковано 8 тра 2019
- This is a talk for people who know code, but who don’t necessarily know machine learning. Learn the ‘new’ paradigm of machine learning, and how models are an alternative implementation for some logic scenarios, as opposed to writing if/then rules and other code. This session will guide you through understanding many of the new concepts in machine learning that you might not be familiar with including eager mode, training loops, optimizers, and loss functions.
Watch more #io19 here: Machine Learning at Google I/O 2019 Playlist → goo.gle/2URpjol
TensorFlow at Google I/O 2019 Playlist → bit.ly/2GW7ZJM
Google I/O 2019 All Sessions Playlist → goo.gle/io19allsessions
Learn more on the I/O Website → google.com/io
Subscribe to the TensorFlow Channel → bit.ly/TensorFlow1
Get started at → www.tensorflow.org/
Speaker(s): Laurence Moroney and Karmel Allison
T700B4 event: Google I/O 2019; re_ty: Publish; product: Cloud - AI and Machine Learning - AI building blocks; fullname: Karmel Allison, Laurence Moroney; - Наука та технологія
I have been studying Deep Learning for the last 3 weeks or so and this guy explained it like in 20 minutes, I wish I had a teacher like him :(
Thanks Akshay! I'm teaching on Coursera if that's any use :)
try siraj? he uses the same methods
@@LaurenceMoroney we are super thankful for your understandable and great explanation 🙏
oh I absolutely agree. There's another fellow on the interwebs that explains it in a similar way but this one is by far the best
@@BrianThomas Thanks! :)
And that’s what an explanation is called.
Wow the way he explained I grasped each and every word of his.
Thanks though
Wow, thanks Raaghav!
Well done Laurence. Your work is inspiring and educative. You are an amazing tutor. I have a big challenge with my Ph.D thesis as it is related with Optimizing Routes using ML and IoT in Waste Management. Is there anyway I can contact you to discuss with you? Via email?
Wow , Thanks Laurence Moroney for making 'Convolutional layer' no more convoluted for me ….also you have made the concept of 'Pooling' so clear .
The best explanation of convolution in few minutes.
🙏
Thanks, Asif!
Shanti bro.
What do you mean by convolution?
I learned it was a mathematical term equivalent to shifting one function along an axis, integrating, and presenting the result of that integration as the 'convolution'. We performed convolution by hand. Then we Fourier transformed the two functions, multiplied them together, Fourier transformed the result back into a real function, and realized that was the same as convolution. It appears you have a different definition.
i think it's a little bit irritating. you have oldschool imageprocessing-convolution which he presents and you can use as preprocessing for a Deep-Learning-MLP. But there are also Neuronal-Networks (CNNs) which are doing similar things by design of the network, and that is more complicated to explain.
Wow I've been reading about this stuff a lot but always had difficulty wrapping my head around it. I did the tensorflow demos, but was unable to understand it well enough to try something on my own. He explained it very clearly and cleaned up a lot of my misunderstanding. Thank you so much for sharing!!!
That's great to hear! Thanks! :)
I can't even tell you how much more I enjoy using TensorFlow 2.0 than I did using the previous versions. Thank you for all the great improvements. A special thank you for TensorBoard. A very useful and fun tool.
Thanks, Bianca!
This is an amazing smooth intro to a relatively complicated topic. The QR codes are a smart move. Hoping to see more videos on similar topics. Thank you both.
Thanks, Eddie!
This is one of the best video on machine learning
Thank you, Balaji!
I just love Laurence Moroneys presentation style. Outstanding ability to read his audience and adjust on the fly in tonality & emphasis. Awesome skill & great presentation
Thanks Dino! :)
Laurence's explanations were wonderful! Thorough, but also simple enough that even a newbie can understand it. Thank you!
Welcome! Glad you enjoyed :)
There should be a button of "Mega Like" for this video. Great explanation of an epic tool. Thanks a lot!
Oh thank you so much! :)
@@laurencemoroney655 thhb4uihh i
Banii=mony ? Cat cost price fel fe recjama say indemnn de a access inainte say apjicatie sa for explcata mai intai cat Costa!!..
@@laurencemoroney655 Great presentation!
It takes a lot of hard thinking to make a topic like this appear this simple.
The Feynman method
Haha. I came up with this talk in about 2 hours in a coffee shop in Tokyo. Maybe their coffee is really really good! :)
@@LaurenceMoroneyThank you for the incredible presentation Mr. Moroney.
It took you just 2 hours you said. I assume it's a similar story as the guy that once charged his client 1000 dollars for a 10minutes service. The client complaint about why he pays so much for only 10 minutes and the guy answered it took him 20 years of experience to do that in 10 minutes.
its simply awesome, i been using jupyter note book, and it has cleared my many queries running in my head, and lot more still remaining,. Good work.
Nice explanation. Thoroughly enjoyed the presentation :)
This is the best video I have seen to demystify TensoFlow, convolution and pooling. Thanks Laurence!
Welcome!
This guy is sssssooooo highly skilled in explaining his topic - its just amazing.
Thank you for that.
Thanks!
This is the best tutorial/introduction that I have ever watched
Oh wow! Thank you :)
(Aside from the talk being great) The presentation screen is awesomely beautiful!! Imagine seeing this for the first time, even just from year 2000.
I know! I had to keep turning around to look at it. Even up close it was beautiful! :)
I need to see it too for real! ^ - ^
I just came here because I today saw the Colaboratorya and then started watching the video. Know it's really helpful for me to understand how easy is TensorFlow and I'm on the right path to learn the all. Thanks to UA-cam and I/O.
Welcome!
I desperately needed an overview like this, thank you!
Welcome!
I cannot believe how understandable you made such a dense topic (mostly) easy to conceptualize! This was wicked! Nicely done!
Thanks, Alex!
@@LaurenceMoroney other way around! I can't wait to go through the rest of the catalog!:)
شكرا على الترجمة إلى اللغة العربية بالسرعة االفائقة و شكرا أيضا على مشاركة هذه القناة عبر الإيميل ونحن نتابع باهتمام كبير التوضيحات والشروحات المتعلقة بأخبار الذكاء الإصطناعي وآخر التطورات في هذا المجال
The best explanation of convolution in few minutes.
🙏
This is amazing!!!!
Thanks! :)
Wow .. this was an awesome presentation, I'll focus on the first speaker since I was somewhat familiar with the subject (I also liked the second one).
He explained all of the key concepts as clearly as I have seen - and I have looked at a LOT of videos, this is a master class in how to do it. I hope he presents on many more related topics. Definitely worth watching.
Thank you so much!
THanks esmail! :)
I have zero experience in Machine Learning, but with this explanation, I feel like I can make a rock paper scissors game with just a little more knowledge.
Awesome -- you can do it! :)
Finally, I got this shit. I have been watching various speeches and talks for the past two years and every time I thought I got it, I didn't. Of course, until I stumbled upon this video. Now I know the concept enough that I can explain it to both technical and none technical people. Thanks for the wonderful video.
Welcome. But it aint $#!+
I did the Introduction to Tensorflow course a couple of months ago on Coursera. I got all that revised in 35 mins. Thanks for this great video.
Welcome!
This is one of the best video on machine learning
This is amazing!!!!
This is amazing!!!!
Best video I have ever seen online to understand deep learning 👏🏼👏🏼
Wow...thanks! :)
Amazing video, there's many elements here that are explained so well. Looking forward to doing more stuff with TF in the future! :)
Thanks! Best of luck with it!
I'm at the end of the first part of the talk and I'm speechless, it's like all the theory I've read and have been taught in class finally falls into place, I never understood how neural networks really "captured" the features, thought it was thanks to pure randomness and just some coincidence, and that the only thing that did the actual work was just minimizing the loss function and that was it, the other things were just experimenting, but after the best explanation on convolutions and pooling I've ever came into contact with now I begin to really understand why this works, so much insights, thank you so much, the first part of the talk was pure gold, the second part was good too, the amount of stuff Keras has implemented and the capabilities to expand it is pretty awesome.
You're welcome! Glad you enjoyed! :)
Whoa, thanks for this video! I learned a lot about making a Model, more about layer, convolutions, pooling for compression, input size
Thanks! :D
Welcome! :)
Very fast and clear talk! Love it! Thanks a lot for sharing!
Glad you enjoyed! :)
I have to say I wish I had gone through this video years ago. It's easier to play with blocks and learn what they look like before trying to learn the milling machinery to make your own blocks without. a clue of what a block looks like or why it looks that way.
Thanks! I hope it was useful to you.
I just learnt basics of Python to gradually start delving deeper into machine learning. This awesome explanation has given me a boost. Thanks Laurence !
Great to hear, thanks Yash! :)
Hey harsh were u learn basic of python plz share
I remember when Laurence taught in Coursera "Tensorflow in practice" course how to recognize that same rock, paper and scissors images, and in the way that he explained it was pretty easy to implement. Excellent teacher!
Thanks Sebastian! :)
I would recommend this video over most others related to machine learning. Good form guys keep up the good work.
// congratulations on the new belt!
Thanks. But what belt? :)
I keep rewatching this vid, it's just so good
Thank you! :)
Wow. 35 min and I learned so much about tensorflow !!
Nice! Glad it was useful for you! :)
www.coursera.org/learn/introduction-tensorflow/home/welcome
Thank me Later
I've been waiting and waiting thinking the only way to properly learn machine learning is through university courses. I've been wrong. I appreciate the work you guys are doing to motivate the next generation of programmers to understand that machine learning is something anyone can understand and use as long as you've put the effort in. Thanks guys :)
Welcome! :)
Thank you very much for this outstanding presentation. Congratulations for your teaching style. Finally, I can understand better those difficult concepts that are presented in such clear simple and clever way.
Awesome, thanks for the feedback!
This is great. I wanna start something in ML. Thanks for sharing
Nice :)
I got started with machine learning today. Your videos are so good that even an absolute beginner like me could understand it
Glad you enjoyed! :)
Very nice talk for people who are beginning to get into machine learning. Thank you for the great explanations!
Welcome! :)
Laurence always killing it! :)
Thanks!
Your explanation is really cool. Learned a lot of things in this 30 minutes of video. You explained every thing that required to quick start with machine learning. I would suggest beginner to start learning from this video. Thanks a lot.
You're welcome! Thanks for your feedback :)
This is the best explanation on the internet, Thank you so much for this classy Talk.
Oh wow!! Thank you so much! :)
No words to congratulate him, what a good teacher, and i saw this video at the right time.
Great! Glad you liked! :)
So simply explained, this is amazing, thank you
Thank you! :)
Respected ma’am, I am Polaka Divya Reddy from India, I came across your profile while applying for MITACS internship on Edge Devices. Your work is truly inspiring!! I was really intrigued by your work on software testing and federated learning.
Your work on systematic mapping was very comprehensive and went in depth! Comparing the quality of automated test scripts has always been a challenge. In federated learning, I do feel that there is a need for better development tools so that we can build these FL systems easily, especially for application level support (APIs). I wanted to know if you faced any efficiency issue in edge FL due to computational power of edge nodes? Also, THANK YOU SO MUCH for these videos, it helped me revise concepts better. I really hope to work with you next year!
Thank you for the useful video. I am curious about deep learning and your explanations showed me a lot of useful insights.
Thanks, Devin!
That's the level of simplification we need...🙌🏻 thanks Laurence
:)
The best explanation of convolution in few minutes.
🙏
this is good
Thanks!
Absolutely amazing video!
I love the way they talk. Excellent, I learned exactly what I was looking for))
That's great, thanks Robert!
@@laurencemoroney655 , by the way I love your videos and would highly appreciate if you could make one about recommendation systems based on matrix factorization, in perticular there practical implementation.
@@robertalaverdyan3150 Oh man. I'd need to learn that first :)
Great presentation! Loved how he explained the concept so clearly. Definitely going to try that codelab.
Thanks, Vivek!
I am making my own language which can combine all the language which had been made till now, .... This video is bit helpful for me to add some magic in it 😃😃
Great talk! Helped to clear gaps in the understanding of conv nets
Great, glad you enjoyed :)
Not everyone who knows something is good at explaining it... this lecturer is fantastic!
Thanks, Angelo!
Thanks for the quality presentation! Really helps understand the process behind TF
Great! Glad you enjoyed! :)
AMAZING presentation!
Thanks!
Great video Laurence! So much information condensed in such a short time. I know what happened here, you wrote drown the script of your talk and then used convolutional layers and max pooling to compress it 😂 . By the way, Karmel, you did awesome too explaining all the deployment options.
Haha! The secret is out! :)
Laurence Moroney was
Super presentation! Nice explanation on concepts in Machine Learning using TensorFlow, 'Convolutional layer,' and 'Pooling'.
Thanks Ravji!
great explanation mr Laurence... simplified explanation for complex things ... I have better understanding listen you. thank you
THanks! :)
Thank you! Always eager to learn.
Welcome!
This is phenomenal.
I have never seen an explanation like this before.
Thanks!
This is how a quality teaching and explanation can help everyone to explore in depth rather than the traditional teaching and old school explanation does...
Thanks!
wow this is just amazing. This is the best explanation of deep learning. Thanks much.
Thanks!
Wow. Thank you so much for this!
You're welcome -- thank you for watching :)
Great content for starters to developers !!
had seen your nlp intro vid that was brilliant loved this one to
Awesome, thank you! :)
Stuff like this makes me so happy I'm a cs major
I was a Physics major :)
Very helpful explanation by Laurence! Hope to find more of his videos.
Thanks! There's lots of them in this channel :)
How it works.
This is the best video that explains everything in detail. Thank you for this.
Thanks!
Great stuff. All my questions are answered. Thank you
WElcome!
Thanks for the amazingly clear explanation!
Welcome!
I might have been watched this 1 million times this week :D
A-ha! So that's where all the views are coming from! :)
@@LaurenceMoroney you're sense of humor is commendable.
Simply the best !
Superbly complements Moroney’s book : AI and Machine Learning for Coders
The feedback from this talk inspired it :)
Machine learning Zero to Hero what a captivating title
Thanks Marco! :)
This is the best artificial intelligence video available. Great explanation. Thank you Google Team.
Thanks! :)
what a man !! i am a surgeon and hardly know how to open the computer and i understand what he did say !!
Oh wow! Thanks Waleed!
Was lucky to see Laurence in person when he did a presentation at my school. Thanks Laurence for an interesting and informative presentation! Got me interested in ML
Thanks Tm! Which school was this at?
It was at UH Manoa
@@tmnic6971 Ah yes! What a great night that was! Hope to go back soon :)
How simply and very articulated explanation.
Thanks Salman!
This is a good explanation. Well done.
Thanks! :)
Nice explanations and tools. I think mainly why I am struggling to make use of machine learning is because I am attempting to implement it from scratch.
Good to start with high level stuff like this, and as you get more familiar, you can peel away the layers to optimize
Excellent tutorial and very easy to follow for someone who is beginner.
Thanks!
And somebody please give this man a hand of applause!
But is that hand a rock, a paper or a scissors? :)
@@laurencemoroney655 I maid this comment related to avengers dialogue " And somebody please give this man a shield".
@@ritik84629 Haha! Thanks :)
Excellent video. Title is aptly named. Learned a lot from it. Thanks a lot.
Thanks so much! :)
Awesome presentation! Thanks Google❤️
Glad you enjoyed, Fuga!
I wish I could I attended it live :(, but this video is the best explanation of convolution you'll ever see!
We were the morning after the party the night before! Many people couldn't make it!
for conv2d layers:
num of parameters = (unm of a filter parameters * num of input features + 1(bias) ) * num of output features.
we can consider the number of features of the first convorlutional layer is 3, which are RGB layers of a picture.
param num : 1792 = (9 * 3 + 1) * 64
param num : 36928 = (9 * 64 + 1) * 64
param num : 73856 = (9 * 64 + 1) * 128
param num : 147584 = (9 * 128 + 1) * 128
Thanks!
is this the crispiest hd quality video on youtube? Excellent content too
You can choose video quality up to 1080p 60
This is amazing!!!!
Machine learning Zero to Hero what a captivating title
Thanks, Lindsey! :)
Salute to you Laurence what a great explanation!!
Thanks Girija! :)
Short and clear! Thank u so much
Thank YOU :)
Really wicked of you Mr Laurence Moroney. Thanks for teaching!
You're welcome, Atal!
Wonderfully and beautifully explained introduction video!
Thanks, Derek!
8:58 , crystal-clear explanation of "Neural Network".
Thanks! It's an approximation to help clarify the concept. Hope it helps :)
Motivation for ML for me, Thank you.
Welcome :)
That is one heck of an explanation!! WOW!
Thanks! :)