How to implement fitnet.m in MATLAB R2007?

Hi All

Is there a way to introduce the fitnet function in MATLAB R2007, despite it's introduced in R2010 ?

Answer 1

[Greg Heath]

There may be licensing problems. Otherwise just use NEWFF. It's old but it is pretty good.

Greg

[farzad]

Thank you so much

I am using this :

Fitnetproblem

But I am still getting error when I only replaced the fitnet with NEWFF

This is the error :

Warning: Could not find an exact (case-sensitive) match for 'NEWFF'.
C:\Program Files\MATLAB\R2007a\toolbox\nnet\nnet\nnnetwork\newff.m is a case-insensitive match and will be used instead.
You can improve the performance of your code by using exact
name matches and we therefore recommend that you update your
usage accordingly. Alternatively, you can disable this warning using
warning('off','MATLAB:dispatcher:InexactMatch').
> In GNNW at 55

To create a neural network using a graphical user interface:
1. Use the command "nntool" to open the Neural Network Manager.
2. Click "New Network" to open the dialog for creating networks.
You can then import or define data and train your network in the GUI
or export your network to the command line for training.

net = 

      divideFcn: 'dividetrain'

??? Undefined function or method 'rng' for input arguments of type 'double'.

Error in ==> GNNW at 57
rng(0)

[farzad]

I think the problem is now rng(0)

[farzad]

Any Ideas?

[Greg Heath]

1. Apparently your first problem was using NEWFF instead of newff.

2. There is absolutely no reason why rng(0) should throw an error. It is just an alternate way to initialize the random number generator

>> rng(0), rand

ans =

    0.8147

>> rng('default'), rand

ans =

    0.8147

Hope this helps,

Greg

[Walter Roberson]

rng() was added in R2011a, replacing the use of rand or randn with the 'seed', 'state' or 'twister' inputs