Answered
Logsig activation function in irradiance post-processing
You have overthought the design. You probably only have to change the random initial weights and/or the number of hidden nodes...

10 years ago | 0

| accepted

Answered
why the network performance decreased??
Given a value for the number of hidden nodes, using different random weight initializations AND random weight divisions will yie...

10 years ago | 0

| accepted

Answered
How can i create ensamble neural networks for pattern recognition? (dividing data)
I don't like huge nets when smaller ones can probably get the job done nearly as well. My first effort would be to see which com...

10 years ago | 0

Answered
Testing NARX on data without known target
You cannot test a net without knowing the target. Hope this helps. *Thank you for formally accepting my answer* Greg

10 years ago | 0

Answered
building customer behavior model using neural network
YOU have to determine what variables are reasonable to use for inputs AND what the target output should be for each input. Th...

10 years ago | 0

Answered
activation of different states in neural network
In that case you should use Gaussian hidden nodes help radbas doc radbas so that one or more will recognize only a part...

10 years ago | 0

| accepted

Answered
how to solve overtrained nn with validation stop?
You cannot over-train unless the net is over-fit. The net is over-fit if there are more unknown parameters than there are tr...

10 years ago | 1

Answered
how to represents a function at three variables from data collected on the field?
If the function is not too weird, the regression neural network FITNET should fill the bill. Assuming x,y,z and t are rows of...

10 years ago | 0

Answered
classification using neural network
No one knows. Everyone has to do their own search. However, if you use the c-dimensional {0-1} column unit vectors as c-class ta...

10 years ago | 0

Answered
How do you format data for a narxnet without preparets?
Using preparets on the narxnet documentation example to deduce the answer: Although I used the default input delay ID = 1:2, ...

10 years ago | 0

Answered
How can I get the function y(t) from NARX problem in Neural Network Time Series Tool?
Since your model is not nonlinear, just use NARXNET with no hidden nodes. X = con2seq(giveninput); T = con2seq(giventa...

10 years ago | 0

| accepted

Answered
activation of different states in neural network
That makes no sense to me. Outputs are not activated randomly. For a given net there is a nonrandom relationship between input a...

10 years ago | 0

Answered
reverse engineering a simple neural network
You did not take into account the default normalization of inputs to the range [-1,1] and the unnormalization of the output from...

10 years ago | 0

| accepted

Answered
using 2D data as input for a neural network
0. An input layer node is not a neuron. Only hidden and output layer nodes are neurons. 1. Use the colon operator matrix(:) t...

10 years ago | 0

| accepted

Answered
Simple Linear SVM classification with normalization
The "I"nput matrix of size [ I N ] = size(x) consists of N I-dimensioal input vectors that contain ALL (includes the question ve...

10 years ago | 0

Answered
How can I get the function y(t) from NARX problem in Neural Network Time Series Tool?
See one of my Narx posts in the NEWSGROUP or ANSWERS: greg narnet Hope this helps. *Thank you for formally accepting m...

10 years ago | 0

Answered
Does weights and bias received from ANN are also normalized???
Typically, the NN training algorithms use the MAPMINMAX transformation to obtain input and target variables in [ -1 1 ] and to r...

10 years ago | 1

Answered
Which architecture and settings are the bets to predict the next value of a time series by time delay neural network?
If you only have 1 series, use NARNET help NARNET doc NARNET Search BOTH the NEWSGROUP and ANSWERS using greg NARNE...

10 years ago | 0

| accepted

Answered
Feature selection using neural network
The most important of I inputs in "I"nput matrices with I rows and N columns depends on the "O"utput target matrix with O rows a...

10 years ago | 0

| accepted

Answered
How to view number of iteration in neural network?
If you use the command line window, that and other useful info is contained in the training record tr: [ net tr ] = tra...

10 years ago | 0

| accepted

Answered
I get a "Performance function replaced with squared error performance" warning when trying to set 'crossentropy' as the performance function.
Crossentropy is, theoretically, not appropriate for regression. Classically, it is only used for classification and pattern-...

10 years ago | 3

Answered
Training feedforward neural network
For N I-dimensional input vectors from c classes, the corresponding N target vectors are {0,1} c-dimensional unit vectors: [...

10 years ago | 0

Answered
Narnet Forecast with removedelay and closeloop
Please revisit my calculations of significant lags. 1. abs(autocorrt) should be sorted 2. The significant correlation thre...

10 years ago | 0

| accepted

Answered
Equation From a simple feedforward neural network
You forgot that the net uses the default MAPMINAX to normalize the input and target before training and, then, to denormalize th...

10 years ago | 0

| accepted

Answered
Which one can be better for noisy time series: the Autoregressive model or Non-Linear Autoregressive Neural network?
1. Significant autocorrelation delays were not calculated 2. No reason was given for for H = 5. What is the length of the ser...

10 years ago | 0

| accepted

Answered
Narxnet Multistep forecast within the time series
1. NARXNET and TIMEDELAYNET require: All input delays to be increasing (not necessarily consecutive) and NONNEGATIVE. All in...

10 years ago | 1

| accepted

Answered
Data replication Neural Networks Matlab
1. I don't understand your question. 2. a. OVERFITTING means there are more unknown weights, Nw, than independent training eq...

10 years ago | 0

| accepted

Answered
Does NARX use the current value of the eXogenous input?
The default input delay ID = 1:d, (d=2) does not use the current value. However, if you wish to use the current value, just a...

10 years ago | 0

| accepted

Answered
Neural Nets: gradient descent issue for polynomial functions
Apparently, a target variance of 9e10 is too much for feedforwardnet to handle. Standardization(zero-mean/unit-variance) does th...

10 years ago | 1

| accepted

Answered
Arranging the input and target matrix in Artificial neural network classification problem.
1. Please reformat your post so that the code will run when cut and pasted. 2. Assuming every thing else is OK, you need to h...

10 years ago | 0

| accepted