The term 'transfer function' is inappropriate here. Transfer functions only apply to linear systems. The term 'activation function' is more appropriate for general neural nets.
So although the neural networks are slow learners, they work eventually getting the right answer, right? So if we just vamped up the processing power we could get the neural networks to become fast learners, theoretically speaking, right?
One question though - at 6:21 you seem to adjust the input nodes with the error factor - why should we do that? Input nodes are our training set values, they should not be adjusted - or am I missing anything?
I've seen this type of example a few times but, I don't see examples of it in action. How would you incorporate this into something like a word search, whereby it can eventually distinguishes what phrase to look for in a Doc based on your input. So for example, a sentence contains "Buy your tickets now and get them cheaper" I type in something like "How do I get cheap tickets" its pumps out "by them now". Am I asking too much?
good
sRahsa 2 days ago
cool!
DaurenTV 4 days ago
Thanks for posting bracksux I'm starting to train a neural network classifier with Weka and this helps as it slowly sinks in
brandondweinberg 1 week ago
this is osum :O
Paramith19 2 weeks ago
thanks. :D
amberongmiciah 3 weeks ago
awesome video! now time to work on my neural network project ;)
seldamat 3 weeks ago
Awesome vid....!! I liked the presentation
dandyprips 3 weeks ago
Hey, U also from TAMU?
I wait till the last and I saw ur email.
Nice presentation. I like it.
oasis4margaret 1 month ago
really nice introduction.thanks
rohitarora1981 1 month ago
very good explanation, you make this easier than i thought
araidewa 1 month ago
neat well explained
tkmallik86 1 month ago
This is great, I wish you made a more advanced part.
JustAnotherGuy66 2 months ago
helpful presentation
thanks a lot
scotlandsimon 2 months ago
I hope you got an A+
Axeltreman 3 months ago
Doesn't take node C the sum of A & B instead of the average?
joeriHer 3 months ago
My brain died
IIIGR1Z 3 months ago
Right on!
zeeEclipse 3 months ago
The term 'transfer function' is inappropriate here. Transfer functions only apply to linear systems. The term 'activation function' is more appropriate for general neural nets.
KiretD 3 months ago
So although the neural networks are slow learners, they work eventually getting the right answer, right? So if we just vamped up the processing power we could get the neural networks to become fast learners, theoretically speaking, right?
isaiahgregov 3 months ago
Great video!
One question though - at 6:21 you seem to adjust the input nodes with the error factor - why should we do that? Input nodes are our training set values, they should not be adjusted - or am I missing anything?
andrewkazyrevich 3 months ago
Very, very good presentation! I loved it, man. I am also trying to get the hang of neural nets, and you gave me the first "baby step".
xplinux22 3 months ago
This has been flagged as spam show
I've seen this type of example a few times but, I don't see examples of it in action. How would you incorporate this into something like a word search, whereby it can eventually distinguishes what phrase to look for in a Doc based on your input. So for example, a sentence contains "Buy your tickets now and get them cheaper" I type in something like "How do I get cheap tickets" its pumps out "by them now". Am I asking too much?
rebirth3X 4 months ago
Comment removed
rebirth3X 4 months ago
does anyone know a really good book on Neural Networks? I want to create a small game that uses Neural networks.
logicCplusplus 4 months ago
thanks mate!
d1d44 4 months ago
really good :) can come up with an example though for one iteration
cssprasad 4 months ago
Very good : clear, interesting and with some subtle humor.
mafLT 4 months ago
unable to enjoy bc vol too low
toddmorrow1 5 months ago
Heh, nice. Though you only covered feed forward networks which learn via backpropagation of error.
What about other ones like ART-nets?
Either way, it was pretty fun to watch :)
Kram1032 5 months ago
A very good tutorial
ScottCMe 6 months ago
Excellent tutorial!
Skatertjah 8 months ago
Wow this is so easy to understand! Thanks so much for making this. I have a test soon about this and man is this going to help me.
NinjaBubbles423 2 years ago
Do you think you could make more videos, this is GREAT!!! LOL :)
maticoff 2 years ago
Thank you, I'm glad you liked it :)
bracksux 2 years ago 2
We learned a lot!
drrobobabe 2 years ago
GOOD JOB!
is0041 2 years ago