Scribd
Upload
271 views2 pages

How To Implement A Multilayer Neural Network in Matlab?: Ask Question

The document discusses implementing a multilayer neural network in Matlab. It provides code to load data, create a neural network with one hidden layer of 8 nodes using newff, train the network on the data, and make predictions on new data. The asker wants to use the outputs of the first network as inputs to train a second network, and asks how to do this using built-in functions while maintaining all 41 input columns. The response recommends using two hidden layers in one network instead of separate networks, and specifies the network can have customizable hidden layer sizes and activation functions.

Uploaded by

meazamali
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
271 views2 pages

How To Implement A Multilayer Neural Network in Matlab?: Ask Question

The document discusses implementing a multilayer neural network in Matlab. It provides code to load data, create a neural network with one hidden layer of 8 nodes using newff, train the network on the data, and make predictions on new data. The asker wants to use the outputs of the first network as inputs to train a second network, and asks how to do this using built-in functions while maintaining all 41 input columns. The response recommends using two hidden layers in one network instead of separate networks, and specifies the network can have customizable hidden layer sizes and activation functions.

Uploaded by

meazamali
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

10/29/2018 how to implement a Multilayer Neural Network in Matlab?

- Stack Overflow

how to implement a Multilayer Neural Ask Question

Network in Matlab?

m1= xlsread('NSL_KDD_TRAIN.xlsx','A2:AO67344'); % m1=input


m= xlsread('NSL_KDD_TRAIN.xlsx','AP2:AP67344'); % m=output

net=newff(m1',m',8);
net=train(net,m1',m');
y=sim(net,m1'); %training data output
y3=round(y); % making nonfraction number

n1= xlsread('NSL_KDD_TRAIN.xlsx','A67345:AO79000'); % n1=input


n= xlsread('NSL_KDD_TRAIN.xlsx','AP67345:AP79000'); % n=output

net=newff(n1',n',8);
net=train(net,n1',n');
y1=sim(net,n1'); %training data output
y31=round(y1);

i want to use the outputs of feed


forward neural networks as input for
training another same kind of neural
network. how can i do that using built-
in function?,from the above code i
need to train a new newff where my
input arguments will be y3 and y31 .
in m1 and n1 a have 41 columns
and i need all 41 columns presence in
the new network. how can i do that
using built-in function?

matlab neural-network

edited Feb 7 '15 at 12:01


Cactus
18.4k 8 47 108

asked Feb 6 '15 at 17:12


Noor Ahmed
3 1

1 Answer
Join Stack Overflow to learn, share knowledge, and build your career.

By using our site, you acknowledge that you have read and understand our Cookie Policy, Privacy Policy, and our Terms of
Service. Sign Up
Why not to use two layers and train
https://stackoverflow.com/questions/28371117/how-to-implement-a-multilayer-neural-network-in-matlab 1/2
10/29/2018 how to implement a Multilayer Neural Network in Matlab? - Stack Overflow

just one network if you assume


architecture is the same?

net = newff(input,target,[Layer_1 Laye

You can specify the size of Layer_1


and Layer_2 as much as you want as
well as you can change activation
function between each of the layers.

Tutorial to create custom network


from Mathworks

Hope this will help

edited Feb 7 '15 at 18:47

answered Feb 7 '15 at 1:48


madbitloman
676 1 11 19

to learn, share knowledge, and build your career.

By using our site, you acknowledge that you have read and understand our , , and our
.

https://stackoverflow.com/questions/28371117/how-to-implement-a-multilayer-neural-network-in-matlab 2/2

You might also like