First video I watched from Siraj and I loved it. It is fantastic how someone can teach something that I didn't learn during several classes on university. Siraj, you were born with the gift to share knowledge and wisdom. Keep doing it and surelly you will do your part into making our planet a better place to live.
The kind of passion and love for the subject which is shown in your lectures proves that you are one of the best in the subject .Hats off to you for doing this great help to the whole of humanity.
Hey Siraj! Thanks for the great tutorials. It is a rare breed that has the technical knowledge AND can articulate the concepts well. Keep up the good work!
I am very thankful. I recently start to explore the AI field but the hardest is to find practical tips and good explanations together. This channel feels like having a friend coaching me into the otherwise overwhelming AI world.
Bro a big salute to you. I watched lots of videos related to this type of topic.what they are do is they just show us trailer and say enroll for our course.but you doing it free of cost.so thank you so much bro. And sorry for bad English.hope u understand what i am trying to say.
I really like how you say "we talked about pooling and we talked about activation" and I get a confused look on my face "what's activation?" and then you say "no I mean we talked about pooling and convolution", but now I am hooked "what's activation?" :) I know maybe it wasn't intentional - but still you've got my attention. SO COOL!
MAN, just wow. You are super brilliant. The way you teach is incredible. I haven't grasped anything in such a quick time from anyone. Please continue this good work of yours.
Minor technicality, most mammals are color blind. Primates reacquired color cones to cash in on the fruit plants evolved to attract color seeing birds and bugs. Interestingly birds have 5 types of cones (UV and something else). There is a huge weird looking type of shrimp with SIXTEEN kinds of rhomboid receptors.
Siraj Raval - The Richard Feynman (as a lecturer) of AI and the world's greatest educator. I learn more from one of his videos than from my combined three years at a public university studying computer science.
Jesus fucking christ you are exceptionally clear minded and eloquent. Literally every minute was perfectly done. Head and shoulders above any other AI channel on youtube with Lex Fridman's work being a close second. Thank you for being excellent at your craft, it is inspiring.
that dropout analogy was on point!!! I have to say that your way of explaining is very natural, unlike the conventional "read from a presentation" method. Thanks!
Great video! I highly respect your passion on this subject. I have been working with neural networks for almost 8 years now. I realized that your approach for simplicity greatly increase the interest of general public and for those who are just starting with this subject. I salute your methodology. Some may critic about the accuracy of the explanation, yes but it really doesn't matter for this is intended for general audience who are making sense on this world of CNN (particularly). So in the end you still have done your job and it's the job of those who wants to fully embrace this knowledge to fill those gaps. (Well of course, we can't expect complete explanation in just 1 video. I find it even hard to explain such topic on audience like this.) So thank you for your efforts! Keep it up!
Hi Siraj! You are by far one of, if not the best Artificial Intelligence Teachers that I have ever found. I am deeply inspired by your videos and they are very clear. I am currently 14 years old, am a rising sophomore, and when I grow up, I want to become the director of AI at Google or my own company. Thanks and I hope that as I continue to learn throughout high school, maybe one day I can be like you. Thanks!
I just want to say. I watched this in July and had no idea what was going on. Watching it again after using cnns for my self driving car course and it's extremely helpful. Great work Siraj!
Thank you!! this video assumes a bit of knowledge for the viewer but expands on the basics of neural networks and expanding on it to include a few other concepts. I suggest one neural networks video from 3 blue 1 brown first, then this is the missing link between the more advanced videos out there.
This is freaking cool video. I love how you deciphered all technical terms, how you put them back together to show details as well as the big picture. Thanks for the great work! This helped me a lot.
Siraj, You are awesome for doing this. Also, you are on that path that I aspire to be on. Thank you for being an inspiration. Indian ethnicity, super smart and intelligent, value giver: you are the definition of a true alpha.
Excellent and dense content (as always) about CNN Siraj. Really appreciate it. All your lectures always force my brain to feed forward and back propagate the information. Very inspiring. For every 10 minute of your video takes me 30-40 minutes to capture and understand in details what you are explaining. Thanks man! You are the best!
This s awesome. I now understand why I sometimes have problem explaining stuff. I try to be correct and in that I cloud recover in to complex thoughts which is not needed to be understand the main point. It's much better t be incorrect and simplify as you do. Good work!
This video is awesome! it made me understand ... how the computer progressively sees the image in more and more complexity/abstraction; how the weights are calculated and that the confusing terminology for weights has to stop! Lol
SIRAJ, I finished the english subs in Math of Intelligence 4, but now i can't make the spanish subs because i'm on evaluation at my high school, projects, test and note in class. I haven't time, But on Friday I go on vacation and i'll make the subs of the last video and the next video (Friday). Thanks and Greetings.
I cannot see English subtitles that aren't auto generated... Working on adding English subs for this video though rn, might not have enough time though. :) Good job on all the previous videos too
Chris B and no haha. This vídeos are very difficult to make the subs. The shorts vídeos i made it in 1-2 hours in Englis with the scripts of Siraj, and traslate it how 3-4 hours, imagine this vídeo of 46 minutes :$
You're one of the reasons I started learning Machine Learning, Before that I thought it's a thing of Genius people but you made things easy. And you look handsome. Amazing way of teaching and that intro rap was awesome.
The amount of amazing content on your channel is incredible. Just wanted to say thank you so much, these videos have been a great source of information.
I don't like when people write code like this (at 34:30) : h = f(X); X = h just use: X = f(X) Whenever I see two statements on one line, I think someone tries to hide something from me and 37:33 it was also a useless source of an error (unless there was something hidden) ;-) You have a nice collection of educational videos, I think I might recommend some of them to people who ask me stuff where I don't have time or knowledge to explain it to them. By the way, time is the only limiting factor, knowledge can be acquired but it takes time and our time is limited. Thanks for your time. :-)
I think a more abstract way of describing convolution is "comparing" instead of "combining". Sure the dot product will combine all values of this matrix operation, but we can think of result as a value of how good the two matrices match. I dont know why but this was important for me :D The example of those two buckets of colors mixing into a new one wasn't that explanatory i guess. Altough the example of the flashlight as being the receptive field was a pretty good one ;D Thank you so much Siraj. You're awesome!
great job Siraj. keep the information out there! this reminds me of when i made a CNN in C# with no libraries. quite a pain but a great learning experience
Siraj, Im spanish, and I follow your channel from very long! My english skills are not very good, but I dont know why I understand EVERYTHING you say, even without subtitles! its amazing ^^ Hope some day you come to Spain, you are invited to the Spanish Direction of CIberdefense of the Spanish Airforce where I work, I ow you a coffee :D hahaha! Thanks for your time!
Hey Siraj, Feature map is filter/kernel applied to the previous layer.Feature map is not the filter/kernel/weight itself.Feature map is the Convolved feature..
Gotta upvote this because i was pretty confused at first until I went to read the textbook and realised he was wrong. The feature map is indeed the convolved feature
Siraj, You are awesome... I am learning so much from you... I have been following you and The Math of Intelligence has been awesome for learning AI.... Many thanks for sharing your knowledge... I think humanity needs AI to survive anythng.. cheers to you...
Really cool somebody talks this thoughts. Nice to hear them in terms of neural networks. Thanks for that! Btw, ur so funny sometimes, haha)) But u try hard, so it's super-cool!
Idk whether it's a good advise, according to your viewers, but for me personally, it becomes hard to perceive ur hyper emotional impact and lots of these "markering" jokes. I know it's good for remembering, and to keep the interest, but that's need more of ones who is not interested enough. For me it's overdone in most cases. That's why science came to its classical language - just to be clear and sharp. It's not good in all cases, but that's my 50 cents)) loved it anyway!!
It's great. I'm working with Udacity Deep Learning where Tensor Flow comes in which simplifies the same operations. But your video just is a great refresher. Allows me to think outside of the Tenor FLow way :) Thanks dude.
awesome!!! your english is loud and clean even to foreign people! You should make (put on your to do list) this kind of video to RNN (Recurrent Neural Network) :)
Man newcomer here! I absolutely love your videos! Especially the ones in which you're serious. I like when it's straight to the point! Thanks for making this happen. You're really helping me.
Great work, Siraj. TH-cam seriously lacks high quality content when it comes to trending complex topics, glad you're filling that void. My only request would be that you spend some time playing around with the output. Keep up the good work !!
You are awesome Siraj. Guys like you provides proof that Humanity is present in this world.Just keep going
Thanks Rishabh. Lots of people helped me growing up. It's time for me to pay it forward.
What? As opposed to humanity not existing? Huh? I don't get it.
@@SirajRaval What a great attitude! I love how positive you are. Thanks for saying this, gives me hope.
th-cam.com/video/Mz3Mu9e0qRQ/w-d-xo.html
@@gaurishgangwar Yup... so much for humanity in people right lol. Humans suck. We need to get on this AI stuff soon as possible
This is awesome Siraj! Keep up the good work! We need more high level TH-camr's explaining these complex topics!
Thanks Erik!
First video I watched from Siraj and I loved it. It is fantastic how someone can teach something that I didn't learn during several classes on university.
Siraj, you were born with the gift to share knowledge and wisdom. Keep doing it and surelly you will do your part into making our planet a better place to live.
The kind of passion and love for the subject which is shown in your lectures proves that you are one of the best in the subject .Hats off to you for doing this great help to the whole of humanity.
I was watching a bunch of videos on CNN, and then Siraj came up in the feed, I was all "YES! THANK YOU!" so refreshing. :)
First person to actually code it and explain it so beautifully. Thanks a lot man helping out your fellow programmers a lot!
Hey Siraj! Thanks for the great tutorials. It is a rare breed that has the technical knowledge AND can articulate the concepts well. Keep up the good work!
Saraj's channel is literally the best AI/programming/math channel you can ever find. And i use his videos in my classes that i TA everytime
And the Noble prize in Teaching goes to Mr. Siraj Raval. Mind blowing explanation.
Dude, I like the way you teach complex topics to be more understandable. It really helps me, thanks!
I am very thankful. I recently start to explore the AI field but the hardest is to find practical tips and good explanations together. This channel feels like having a friend coaching me into the otherwise overwhelming AI world.
Bro a big salute to you.
I watched lots of videos related to this type of topic.what they are do is they just show us trailer and say enroll for our course.but you doing it free of cost.so thank you so much bro.
And sorry for bad English.hope u understand what i am trying to say.
The best thing we can say to an engineer about his work is that he do magic, so I say that "this video is magic!" Thank you Siraj!
I really like how you say "we talked about pooling and we talked about activation" and I get a confused look on my face "what's activation?" and then you say "no I mean we talked about pooling and convolution", but now I am hooked "what's activation?" :) I know maybe it wasn't intentional - but still you've got my attention. SO COOL!
This is by far one of the most comprehensive yet brief tutorial on neural networks I have seen. Thanks @Siraj
I really appreciate the way you explained convolution before jumping to the coding part.way to go!
MAN, just wow. You are super brilliant. The way you teach is incredible. I haven't grasped anything in such a quick time from anyone. Please continue this good work of yours.
wow!! i spent the whole week reading "Deep Learning" book. this video just made me amazed. now i've the overall idea. thanks Siraj!!
Minor technicality, most mammals are color blind. Primates reacquired color cones to cash in on the fruit plants evolved to attract color seeing birds and bugs. Interestingly birds have 5 types of cones (UV and something else). There is a huge weird looking type of shrimp with SIXTEEN kinds of rhomboid receptors.
Oh, you mean /this/ shrimp - theoatmeal.com/comics/mantis_shrimp
Of all the tech videos I watch, your's are the only ones I watch at 1x - real time - because you're real! Thanks for doing this!
Siraj Raval - The Richard Feynman (as a lecturer) of AI and the world's greatest educator. I learn more from one of his videos than from my combined three years at a public university studying computer science.
Thanks for giving one of my first projects in Deep learning
I love that novel experience analogy. I see how people might think you are completely on the wrong track or even crazy but I totally get it. Awesome!
I do love this series a lot more than previous videos :), less music and buzz words, more technical stuff and explanations. Really god work.
Jesus fucking christ you are exceptionally clear minded and eloquent. Literally every minute was perfectly done. Head and shoulders above any other AI channel on youtube with Lex Fridman's work being a close second. Thank you for being excellent at your craft, it is inspiring.
that dropout analogy was on point!!! I have to say that your way of explaining is very natural, unlike the conventional "read from a presentation" method. Thanks!
Love you Siraj. One of the TOP teachers on youtube. I don't care what people say.
This is like hitting a pot of gold. your videos are the best I have come across on youtube. Thank you much :)
Great video! I highly respect your passion on this subject.
I have been working with neural networks for almost 8 years now.
I realized that your approach for simplicity greatly increase the interest of general public and for those who are just starting with this subject.
I salute your methodology.
Some may critic about the accuracy of the explanation, yes but it really doesn't matter for this is intended for general audience who are making sense on this world of CNN (particularly). So in the end you still have done your job and it's the job of those who wants to fully embrace this knowledge to fill those gaps. (Well of course, we can't expect complete explanation in just 1 video. I find it even hard to explain such topic on audience like this.)
So thank you for your efforts! Keep it up!
Hi Siraj! You are by far one of, if not the best Artificial Intelligence Teachers that I have ever found. I am deeply inspired by your videos and they are very clear. I am currently 14 years old, am a rising sophomore, and when I grow up, I want to become the director of AI at Google or my own company. Thanks and I hope that as I continue to learn throughout high school, maybe one day I can be like you. Thanks!
hi thank you love u
I just want to say. I watched this in July and had no idea what was going on. Watching it again after using cnns for my self driving car course and it's extremely helpful. Great work Siraj!
great to hear~!
Thank you!! this video assumes a bit of knowledge for the viewer but expands on the basics of neural networks and expanding on it to include a few other concepts.
I suggest one neural networks video from 3 blue 1 brown first, then this is the missing link between the more advanced videos out there.
I am grateful to you for providing such quality videos.
Dang these Indians are smart.. Khan Academy and this Chanel are living proof.
This is freaking cool video. I love how you deciphered all technical terms, how you put them back together to show details as well as the big picture. Thanks for the great work! This helped me a lot.
Siraj, u r super in explaining complex things in simple way. Kudos to great work.
I heard bad stuff about you but your way of explaining things is remarkable.
you are very good men "Siraj" ... "Siraj" it mean in our language (arabic language) it mean (LIGHT) ... so you are the light of our way. thank you
SiraJ - LOVE YOU MAN ! YOU ARE THE MAN, reminds of my younger day.. about AI, nobody understood it 25 years back during my graduate studies.
Yo Siraj, you are straight up turning into that cool teacher everyone wants in high school
I never really comment on youtube but (10000*thumbs) up Siraj, thanks for this in-depth explanation of what we usually call "the black box".
Siraj, You are awesome for doing this. Also, you are on that path that I aspire to be on. Thank you for being an inspiration. Indian ethnicity, super smart and intelligent, value giver: you are the definition of a true alpha.
Siraj, I love your videos! Because you're a brilliant guy who just loves to teach. And for me, there's no better way to learn!
Excellent and dense content (as always) about CNN Siraj. Really appreciate it. All your lectures always force my brain to feed forward and back propagate the information. Very inspiring. For every 10 minute of your video takes me 30-40 minutes to capture and understand in details what you are explaining. Thanks man! You are the best!
This is easy to follow, very well and intuitively explained, thanks Siraj
This s awesome. I now understand why I sometimes have problem explaining stuff. I try to be correct and in that I cloud recover in to complex thoughts which is not needed to be understand the main point. It's much better t be incorrect and simplify as you do. Good work!
This video is awesome! it made me understand ... how the computer progressively sees the image in more and more complexity/abstraction; how the weights are calculated and that the confusing terminology for weights has to stop! Lol
SIRAJ, I finished the english subs in Math of Intelligence 4, but now i can't make the spanish subs because i'm on evaluation at my high school, projects, test and note in class.
I haven't time, But on Friday I go on vacation and i'll make the subs of the last video and the next video (Friday). Thanks and Greetings.
I cannot see English subtitles that aren't auto generated... Working on adding English subs for this video though rn, might not have enough time though. :) Good job on all the previous videos too
Chris B because Siraj still is not accept it, i don't know why.
Thx so much Chris.
Chris B and no haha.
This vídeos are very difficult to make the subs. The shorts vídeos i made it in 1-2 hours in Englis with the scripts of Siraj, and traslate it how 3-4 hours, imagine this vídeo of 46 minutes :$
Pinkie The Smiling Cat Yeah, I can contribute to the English CC thought a bit at a time...
accepted the submissions, thanks so much Pinkie!
Best video i got on TH-cam about machine learning
Very high clarity and involved explanation. It is one of the best session. Thanks for your efforts.
I love how you used psychedelics and slydiving as a metaphor for Dropout. Turn on, tune in, *drop out*; or literally dropping out of a plane... Haha
You just convinced me to take psychedelics
make sure to do with with a friend u trust
siraj knows
It IS excellent advice in general, though Jordan was joking.
i love that siraj can give good advice on anything :)
John Smith 26:30
This seriously needs and deserves a lot more views!
Thanks Siraj! Keep up the good work.
The best explanation on TH-cam about CNN. At 7:21 I cried!
Nice to see the implementation is straight numpy - thanks!
This was so awesome...I now understand CNN completely.... thank you so very much.
8:13 Finally ! After so much elementary blah blah, finally one useful information (a little math, finally). Thank.
This is one of the best videos I've watched on neural networks. I'm in love with the concept! Thanks a lot Siraj :)
You're one of the reasons I started learning Machine Learning, Before that I thought it's a thing of Genius people but you made things easy. And you look handsome. Amazing way of teaching and that intro rap was awesome.
Serously...
You are amazing I ever found on youtube.
we are thankful to you for making such nice deep learning tutorials with fundamentals
The amount of amazing content on your channel is incredible. Just wanted to say thank you so much, these videos have been a great source of information.
you are the best explainer ever i have seen
1:32 am ok thats enough youtube for toni- wait what oh siraj uploaded today... rip sleep
hi Spark
I only need one word to explain this video! Amazing!!!!!
7:36
Siraj, you are the man. Thanks for the love and right back at you my friend!
Thanks Mark!
I don't like when people write code like this (at 34:30) :
h = f(X); X = h
just use:
X = f(X)
Whenever I see two statements on one line, I think someone tries to hide something from me and 37:33 it was also a useless source of an error (unless there was something hidden) ;-)
You have a nice collection of educational videos, I think I might recommend some of them to people who ask me stuff where I don't have time or knowledge to explain it to them. By the way, time is the only limiting factor, knowledge can be acquired but it takes time and our time is limited. Thanks for your time. :-)
I subscribed the very moment you said - we are awesome. We are programmers.
Great Siraj. Keep up the great work.
I think a more abstract way of describing convolution is "comparing" instead of "combining". Sure the dot product will combine all values of this matrix operation, but we can think of result as a value of how good the two matrices match.
I dont know why but this was important for me :D The example of those two buckets of colors mixing into a new one wasn't that explanatory i guess. Altough the example of the flashlight as being the receptive field was a pretty good one ;D
Thank you so much Siraj. You're awesome!
You are the most beautiful human being ever created .Greetings from Argentina
great job Siraj. keep the information out there! this reminds me of when i made a CNN in C# with no libraries. quite a pain but a great learning experience
this guy is amazing. i wish u were my university of all subjects at this time I would be doing crazy things
Siraj, Im spanish, and I follow your channel from very long! My english skills are not very good, but I dont know why I understand EVERYTHING you say, even without subtitles! its amazing ^^ Hope some day you come to Spain, you are invited to the Spanish Direction of CIberdefense of the Spanish Airforce where I work, I ow you a coffee :D hahaha! Thanks for your time!
Lol I cannot help pressing the like button! Thanks for your compliment and you are super coooool as your subscribers are ;)
Hey Siraj, Feature map is filter/kernel applied to the previous layer.Feature map is not the filter/kernel/weight itself.Feature map is the Convolved feature..
THat's what I thought as well.... It's the result of the filter/convolution-operation ... right? Just to confirm....
@@alfianabdulhalin1873 exactly
Gotta upvote this because i was pretty confused at first until I went to read the textbook and realised he was wrong. The feature map is indeed the convolved feature
@@puttatidam.1819 what is the book you referred if you please can tell? i will be obliged.
little mistake from him. you can use the deep learning book of Bengio MIT edition
Thank you for explaining it in a very easy way
Great (and funny) explanation of the math and the code. Nice comparison between drop out procedure and old people :)
Thanks for doing this. Your videos inspired me to start learning about neural nets.
Dropout and over fitting explanation was awesome at 25:30 .. Thanks Siraj . It was really helpful to understand
"It's beautiful, you're beautiful" LMAO, too cool, Siraj. Too cool. Keep up the awesome work and the huge flow of positivity + knowledge! :D
cool channel .......I've been wanting to update my computer science ......superb content........& a rad doo !
You are awesome... I am awesome too.No one knew we existed...But you are way cooler...I like that part of this video
I am his fan bro.. I am learning much from him.
yikes.. i cant believe i understood in 30 min what i strugled with in a semester!.. tnx :D
Siraj, You are awesome... I am learning so much from you... I have been following you and The Math of Intelligence has been awesome for learning AI.... Many thanks for sharing your knowledge... I think humanity needs AI to survive anythng.. cheers to you...
Great video! Love getting to learn ML at my own pace and you provide so much detail!
you should make a video just of the part "we are smart and we are cool"
this saturday
awsome
That was my favorite part of the video.
good idea
Agree, then we can have a more concise video without those meaningless bluffing. And another video of the achievement and how great this is.
Great Explanation!!! Went through so many videos but none convinced me as this one
Thank you very much for your help and the time dedicated to make this video Siraj! Regards from Perú.
Really cool somebody talks this thoughts. Nice to hear them in terms of neural networks. Thanks for that! Btw, ur so funny sometimes, haha)) But u try hard, so it's super-cool!
Idk whether it's a good advise, according to your viewers, but for me personally, it becomes hard to perceive ur hyper emotional impact and lots of these "markering" jokes. I know it's good for remembering, and to keep the interest, but that's need more of ones who is not interested enough. For me it's overdone in most cases. That's why science came to its classical language - just to be clear and sharp. It's not good in all cases, but that's my 50 cents)) loved it anyway!!
Siraj, great tutorial! I like how passionate and excited you are when talking about CNN!
cleared all of my problems with cnn, Thanks
First video explained convolution so clearly. Great job. man.
It's great. I'm working with Udacity Deep Learning where Tensor Flow comes in which simplifies the same operations. But your video just is a great refresher. Allows me to think outside of the Tenor FLow way :) Thanks dude.
awesome!!! your english is loud and clean even to foreign people! You should make (put on your to do list) this kind of video to RNN (Recurrent Neural Network) :)
coming
I love it when u make a neural net from scratch (numpy)
we are smart and we are cool ! But you are amazing Siraj ! thanks for all this cool stuff you make to help all of us
I love your energy, thanks for sharing this!
Man, you banging with those vids. Keep em coming, the quality and informativeness and clearness never cease to amaze :)
Super intuitive. Just. Awesome.
Man newcomer here! I absolutely love your videos! Especially the ones in which you're serious. I like when it's straight to the point!
Thanks for making this happen. You're really helping me.
Great work, Siraj. TH-cam seriously lacks high quality content when it comes to trending complex topics, glad you're filling that void. My only request would be that you spend some time playing around with the output. Keep up the good work !!