Lecture - 36 Leavning Using neural Networks - I

[ptrigueiros67]

Very well presented. The English is very good, so we can understand very well the lecture. There are some teachers that impossible to understand.... It was very helpful. Thanks

Very beautiful teacher :-)

[aperezNWO]

Very Thanks for Sharing

[vxgokulvx]

at 57:20.

Actually the sigmoid function becomes steeper for k > 1. The graph show here is for k = 10. NOT k= 0.1

[vxgokulvx]

Nice lecture! but some minor errors.

at 22:50, the statement made is "delta w i is +ve when X i is +ve and delta w i -s -ve when x i is -ve."

I guess, this is wrong. The sign of the delta w is not dependent on x i. rather, it depends on the sign of the error. i.e (t - o). if (t - o) is -ve, reduce W. Else if (t-o) is +ve: increase W. Else if (t-o) is zero: no change.

[manavTHEDARKKNIGHT]

man,don't teach unless you have any idea about the subject................very bad video.....

[TheReasonWhyGuy]

@manavTHEDARKKNIGHT This must either be extremely basic stuff for you, or it's far ahead of your experience, and you decided to attack the teacher for your not understanding.

Either way, everything in this video checks out with what I know about neural networks. :P

If this is very basic stuff to you, then don't criticize the teacher... that would be like going into a kindergarten class and criticizing the teacher for only teaching the basics, and not teaching algebra or some shit like that :P

[mars2k07]

excellent video and yes the professor is certainly a radiant and captivating presence. I feel like I'm trying harder than usual to understand because I don't want the good lady to think I'm thick!. Mind you, I don't think this message is going to impress anybody ..

[vangtid]

great and very beautiful teacher ! cheer