Respondida
Closedloop Recurrent Neural Network problem
I know this is not a good answer, but I and others are also having closeloop performance problems. So far I haven't found out wh...

más de 13 años hace | 0

Respondida
How to further debug this code?
1. If the number of unknowns is greater than the number of equations, then a solution is not unique (How many solutions to x1 ...

más de 13 años hace | 2

| aceptada

Respondida
Problem with network obtain lagged values NarX
Please run your code on a nndataset so we can compare.

más de 13 años hace | 0

| aceptada

Respondida
can images be given as inputs to neural network GUI toolbox?
1. Use feature extraction to reduce the number of pixels 2. Columnize the resulting image into a vector using (:) Hope thi...

más de 13 años hace | 0

| aceptada

Respondida
serial classification of timeseries "phases" with neural network
Inputs to both both static and dynamic nets have a fixed number of inputs. There may be problems if the significant temporal len...

más de 13 años hace | 0

| aceptada

Respondida
Passing New Data to Neural Net
Upon rereading your post, I realized that this is NOT A TIMESERIES PROBLEM UNLESS your successive inputs are serially correlate...

más de 13 años hace | 0

| aceptada

Respondida
This is a question about time series forecasting
1. Why are you using elmannet instead of narnet? Didn't you read the warning and recommendation in the elmannet documentation? ...

más de 13 años hace | 0

| aceptada

Respondida
Passing New Data to Neural Net
Which of the nndatasets should we use with the code you posted? help nndatasets

más de 13 años hace | 0

Respondida
Delayed outputs with respect to the original NarX
Please be more specific. Use one of MATLABs nndatasets and post relevant code.

más de 13 años hace | 0

Respondida
serial classification of timeseries "phases" with neural network
1. Obtain N training pairs of I-dimensional inputs and corresponding O-dimensional targets 2. If the task is classification b...

más de 13 años hace | 0

Respondida
How can i optimise weights for BPN using genetic algorithm?
The value MSE = 0.0054 means absolutely nothing to me because I don't know the scale of the target data. On the other hand, ...

más de 13 años hace | 0

| aceptada

Respondida
How can i optimise weights for BPN using genetic algorithm?
I have read that ... Where? ... The internet? Please clarify. Is your prediction problem static (simultaneous input and outpu...

más de 13 años hace | 0

Respondida
How can I simulate a NARX neural network after being trained and tested?
% How can I simulate a NARX neural network after being trained and tested? 1. You have not suggested a MATLAB nndataset to us...

más de 13 años hace | 2

| aceptada

Respondida
Problem with network NarX need help for final step
If you look at the significant lags of the output autocorrelation function and the output/input crosscorrelation function, you c...

más de 13 años hace | 0

| aceptada

Respondida
how to do comparison on neural network output
>> net1 = fitnet;net2=fitnet;isequal(net1,net2) ans = 1 Hope this helps. *Thank you for formally accepting my a...

más de 13 años hace | 1

| aceptada

Respondida
how to get the Hidden layer output???
y = b2 + LW*h % PURELIN OUTPUT h = LW\(y-b2) % Hidden Layer output However, the solution is not unique if null(LW) is...

más de 13 años hace | 0

| aceptada

Respondida
Problem with network NarX need help for final step
If you want to predict exactly 10 days ahead, if possible, design a net that will do that. Iterating a net that will predict ...

más de 13 años hace | 0

Respondida
save the neural network output "net" to matfile
I don't know why you are using the term 'features' for describing a net. Nets can be thought of as structures with different siz...

más de 13 años hace | 1

| aceptada

Respondida
Neural Network NAR-based time-series prediction starts failing after several timesteps
I took a look at the openloop problem N = 1000 dx = 0.1 x = dx*(0:N-1); t = cos(x); [ I N ] = size(x) [O N ] =...

más de 13 años hace | 0

Respondida
Multi Input Neural Network with delayed target input.
To be consistent with MATLAB notation, rename the target t and the output y. What are the significant correlation delays in t...

más de 13 años hace | 0

| aceptada

Respondida
Neural Network NAR-based time-series prediction starts failing after several timesteps
After a not so quick look I have the following comments: 1. cos(t) has a period of 2*pi and satisfies a 2nd order homogeneous...

más de 13 años hace | 1

| aceptada

Respondida
problem with closeloop narx
You can iterate as long as you have exogeneous inputs. If your last exogeneous input value is at t = tf, your last predicted out...

más de 13 años hace | 0

Respondida
Change Transfer Function on a Neural Network
The training logic for NEWRB is quite different from the other nets. I doubt very much if it is possible to do what you propose....

más de 13 años hace | 0

| aceptada

Respondida
RBF Neural networks
http://www.mathworks.com/matlabcentral/newsreader/view_thread/151286#888395

más de 13 años hace | 1

| aceptada

Respondida
How can I select the centers of the first hidden layer of nn with the orthogonal least squares algorithm????
MATLAB's NEWRB neither 1. Allows a specified starting configuration 2. Allows RBFs with different widths. For either, ...

más de 13 años hace | 0

| aceptada

Respondida
Problem in training the feed-forward network.
load input load output p=input'; t=output'; [ I N } = size(p) % [ 1 3988 ] [ O N } = size(t) % [ 1 3988 ]...

más de 13 años hace | 0

| aceptada

Respondida
labeling target in neural network in matlab
Your basic problem is input variable subset selection. Say from 600 to 39. However, if you design 40 classifiers, you have 40 d...

más de 13 años hace | 0

Respondida
How to write custom performance function for neural network in MATLAB 2012b (neural network toolbox 8.0)
I think we should start where you left off. So, post your code. With relevant comments. Which entropy function? Exclusive c...

más de 13 años hace | 0

| aceptada

Respondida
labeling target in neural network in matlab
You have 40 subjects. Each subject is a class. You have 5 I0-dimensional input vectors (I0 = 600) for each class. 1. If the ...

más de 13 años hace | 0

| aceptada

Respondida
input of neural network
If you have N I-dimensional input vectors and the corresponding N O-dimensional output vectors, the input and output matrices ...

más de 13 años hace | 0

| aceptada