Lecture - 27 Learning : Neural Networks

[nightowl8936]

This video is average quality. The relatively archaic explanation methods and strong mathematical bias place burdens on the learner to be good at math and able to follow the topics in an order thats not necessarily the easiest.

[venkatarun95]

@nightowl8936 IITan's are expected to know Maths!!

[nightowl8936]

@venkatarun95

Typical 3rd world mentality. Middle east schools are similar in that they focus only on the harder sciences and stupidly think those are the only ones that matter.

Indians are forgetting their intellectual history of a broad spectrum of ideas.

To properly understand learning in general, you need to understand psychology, and neuroscience, and cognitive science before you can fully grasp backpropagation being just one way of doing it.

You know you're not that bright.

[raptor12143]

Seems like that the professor is very confused... He's also skipping topics by making excuse "i will tell in a mo" n then forget. Not good for newbies :( . One must read book to learn basics before watching this video.

[quantamflux222]

I took SYDE 558 at UW. It is a fairly conceptual course, and I didnt apply it to my design project. It was overall made of good learning material.

[ahmer0623]

i dont know why these university professors are so unprofessional and lazy....

[robextra0]

I used to teach that 20 years ago, but with a more visual method, using color interactive graphics (not as developed as now, but still quite helpful) to understand the different steps and show the convergence over time.

regarding BIAS: it does NOT matter whether you use + or -, as long as you are consistent throughout the learning and operation phase.

[SalsaTiger83]

we can look forward to the khanacademy version of this ;-)

[thinkiingfellow]

Gr8 video and a useful lec material

[sacaafrica]

I cant understand onething: in some books I find that in= sum wa+ Bias and in other ones in = sum wa- bias !!! wich of them is the coorect?

please help

[theoark]

@sacaafrica consider the case of bias being negative.

[sacaafrica]

@theoark according to the famous book " Neural Networks A Comprehensive Foundation" in= W A+ BIAS. other books say that threshold in biological neurons has a negative value but in the artificial it may be positive, in this case, it is usally reffered as bias. and that what makes me confused!!!! if you have any explanation you can help me.

Thank you for responding

[Xanderiel]

why does he talk like Apu?

[dragosboros]

another good talk

[selotmani1]

thank you for sharing

[gaBehcuoDsuoitneterP]

A node is like a neuron? But a neuron's move and a node does not?

[folatube]

Good learning material.

Its great to see more of this than mindless "entertainment" on youtube.

[vickytrivedi]

He did a great work. It is an amazing source of knowledge. Stop commenting on trivial stuff (his writing or paper sheets) and please please don't abuse the right of commenting by doing this.

[exc390]

He meant mean squared error in 24:10 rather than root mean squared...

[Dollarbob]

duh

[hadeedraheel]

how u have added a hyperlink at 24.10

[vinhbt123456]

Very interesting and informative lecture. Thanks million

[shoaibjameel123]

Great lecture but you waste a lot of paper. Why not use even the back side of the paper?

[asldfkjgl]

you're an idiot hes using marker it would bleed through

[shoaibjameel123]

Yes, I am an idiot!

My point is to let him know that he is wasting paper and think of some alternative ways to save paper. No wonder our world is doomed because of such mindset.

[asldfkjgl]

Hows this a waste? That is what paper is intended to do. He can't use the back since he is using marker and it would show through. Also these drawings are being used by thousands of people who have now watched the video. So each sheet of paper is useful for thousands of people. Now think about a student in this guys class taking down notes, only that student will use his or her notes. So in essense each of these pages is orders of magnitude more useful than students taking notes. stop them

[shoaibjameel123]

You may be affirmative.

I've heard that Mahatma Gandhi used to write on the free space of letters which used to come to him. Imagine how dirty those letters would have been since they used to come from afar and Mahatma still used to write on them. That's why he is the Mahatma and is still respected.

Nevermind, I will not get into more arguments but ways have to be found out to save our environment as best as possible. Even one paper is important.

[whistlindude]

very helpful but a bit slow .........

thnx sir

[bubach85]

Whats up with that freaky looking fingernail? Why save just one to that lenght? :S

[carpetmermaid]

it is a nose cleaner/picker for himself and/or his kids...lots of people use their pinky nail for that purpose.

[moseskiiza]

@bubach85  lol, man you idiot, he probably plays a string instrument (guitar or sitara), i know you know that ppreciate your ahumour tho

[gaBehcuoDsuoitneterP]

@moseskiiza this guy is a musician fooling everybody.

[999madcows]

Thank you so much for posting this lecture! I'm a student very interesting in learning about neural networks and this video has taught me a lot. I've noticed that most of the diagrams are from "Artificial Intelligence A Modern Approach - by Stuart J. Russell , Peter Norvig" and that is the same book I am reading!

Thanks again!

[candoyo]

Amazing video ... great examples.. love NN!!!... saved me a lot of time reading the lecture slides ;) way to go IIT!!!

[palash291]

Nice Tutorial!!....Its very great to see our IITs coming out with such good initiatives in the times when engineering youth do nothing except jerking all the time :P ....Prof Dasgupta, though a bit confused, really rocks in the video!!....In fact, IIT profs are rocking in their own way!!

[benadam777]

Excellent, excellent training video. One thing I noticed however, is that the error function is demonstrating method of least squares and not root mean square, if I'm not mistaken. Thank you for the very informative lecture.

[smaqsood123]

you are right..

Please visit this lecture from MIT for visualization of Least Squre Mathod..

Lec 21 | MIT 18.086 Mathematical Methods for Engineers II

[veronicle]

Great...thankas

[vangtid]

finally a great tutorial on NN

[Norman60Fahrer]

I very much like these lectures from India Universities.

Interesting to see how students got confused when he did calculations for error-minimizatin. Using x as multiplication sign and as a variable is certainly not helpful in writing equations.

[bayrees]

@Norman60Fahrer

How many times did (do) you confuse between gas pedal and break pedal when you are driving @ high speeds?

[bollamrahul]

nice lecture i think this can bang the illiterates of this subject in teaching profession.

[h0mee]

you rock! great walk through of challenges faced by perceptron

[neoniator]

CHeck your system sound

[shriharireddy]

can i get next lecture of this

[csnmgeek]

thanks Dr. (y)

[melhosseini]

Professor Dasgupta is one of the scientist any country dream of him,

thank you professor for this remarkable lectures, and thank you You Tube

N.B: how can i have a copy of this Great lecture, it will ease teaching my student AI

can anyone help me

[MartinMusic1]

I don't know how about the copy rights, but there is a program to easy download a youtube film to computer. It's called 'save2pc' and can be found through google.

[anukefromrussia]

This was so great! I now understand neural networks!!!

[nightowl8936]

The mathematics of this lecture are way above the average YouTube user.

The typical YouTuber is interested in girls shaking their asses around. Those are the vids that get millions of views.

But for those that care about machine learning, this is roughly equivalent to an upper year computer science lecture.

Valuable academic content in this video, but could have been more professionally presented and edited.

4 stars.