Got two books on Neural Nets, neither cover noting more than 1 hidden layer.

Oh, anyone recombed a book, really would like one that covered two or more hidden layers..

Author

Message

Got two books on Neural Nets, neither cover noting more than 1 hidden layer.

Oh, anyone recombed a book, really would like one that covered two or more hidden layers..

This topic has a solution. Jump to the solution.

Last Edited: Sat. May 9, 2020 - 09:50 PM

Anyone here good with Neural Nets or equally really good at maths. Have a look at the attached document with the full question listed!

try this...has code you can try & you can instal Python for free (get Python at https://www.python.org/)

https://machinelearningmastery.com/implement-backpropagation-algorithm-scratch-python/

Another decent read:

https://medium.com/datathings/neural-networks-and-backpropagation-explained-in-a-simple-way-f540a3611f5e

I somewhat remember a course I took around 1993 using my 100 MHz PC to train on Lenna:

https://en.wikipedia.org/wiki/Lenna

Last Edited: Sun. Apr 19, 2020 - 03:59 PM

Hello freaks. So I was wondering if you could help me? I'm looking to know how to calculate a 2 hidden layer network. Have a look at the diagram below.

To calculate w1 I have the following.

The calculation of the first term on the right hand side of the equation above is a bit more involved than previous calculations since affects the error through both and .

Now my question is. If I had a hypothetical weight input into X1 called weight I, wi what is the equation for back propagating tht weight? I'm thinking it is maybe it is this.

âˆ‚E/âˆ‚wi = âˆ‚E/âˆ‚x1 . âˆ‚x1/âˆ‚zx1 . âˆ‚zx1/âˆ‚w1

Therefore:

** âˆ‚E/âˆ‚x1 = ( âˆ‚E/âˆ‚o1 . âˆ‚o1/âˆ‚zo1 . âˆ‚zo1/âˆ‚x1) + (âˆ‚E/âˆ‚o2 . âˆ‚o2/âˆ‚zo2 . âˆ‚zo2/âˆ‚x1)**

Or maybe it is this:

** ****âˆ‚E/âˆ‚x1 = ( âˆ‚E/âˆ‚h1 . âˆ‚h1/âˆ‚zh1 . âˆ‚zh1/âˆ‚x1) + (âˆ‚E/âˆ‚h2 . âˆ‚h2/âˆ‚zh2 . âˆ‚zh2/âˆ‚x1)**

I really hope you can help me with this, thanks for having a look!

Wm.

Ah, I didn't see this thread before I replied to your other one here:

https://www.avrfreaks.net/comment/2896386#comment-2896386

... so never mind ;-)

This reply has been marked as the solution. #6

This function does it, just substitute for db1 for the weights.