Answered
Performance in Neural Network Training
The main function of the validation subset is to stop training if it's error rate increases continually for 6(default) straight ...

9 years ago | 0

Answered
Performance comparison plotting for different Back propagation algorithms
Your stopping criterion for regression should be a fraction of the mean target variance. I tend to choose 0.01 (Rsquare = 0.99) ...

9 years ago | 0

Answered
How to implement feature selection on a fitting ANN?
Although there are fancy ways to rank variables for neural networks, most are (in my opinion, too complicated to waste time over...

9 years ago | 0

Answered
Need help about ANN output- Can't get Output as 0 and 1 but between 0 and 1
Classification target vector components should be nonnegative conditional (on the input) prior probabilities that sum to 1. ...

9 years ago | 0

Answered
Selection of Neural Network Training Data
Before learning, obtain the mean and standard deviations of the input and target variables. Overlay the plots of the variables o...

9 years ago | 0

Answered
Substitute of CUSUM algorithm for point-change detection and curve fitting..
Just compare the new point with the old mean before updating the mean check(n+1) = abs( x(n+1)- meanx(n) ) meanx(n+1)...

9 years ago | 0

| accepted

Answered
NaN performance and gradient in RNN
Your configuration doesn't make sense to me. If you want to classify a series 1. Find the significant correlation lengths of ...

9 years ago | 0

Answered
It's possible to make forecasting with Neural Net Clustering ?
I would not use clustering for prediction. However, you could use NEWRB. Nevertheless, I think your best bet is to use the NARX...

9 years ago | 0

| accepted

Answered
Dropout Neural Network for Classification
Drop out is explained in the MATLAB website. Input values are randomly set to 0.0 with a specified probability. This is a mitiga...

9 years ago | 0

Answered
How to change the performance function of neural network to mean absolute relative error
You have at least 2 obstacles: 1. When t --> 0 2. abs is not differentiable If t --> 0 is not a problem try myp...

9 years ago | 0

Answered
Custom Neural Netwrok (Manually re-implementing "patternnet' using "network")
1. Initialize the RNG to the same initial state. 2. Transform inputs and targets to [-1 1 ] before learning and transform the...

9 years ago | 0

Answered
Getting an error while using a user defined transfer function in newff function
tansig(z) = tanh(z) Saves a lot of headaches! Greg

9 years ago | 0

| accepted

Answered
[net,tr,Y,E]=train(net,P,T); How Y and E get their dimensions with this syntax?
For pattern recognition use NEWPR which is the special case of NEWFF that is designed for classification and pattern recognition...

9 years ago | 0

Answered
What is the MATLAB code to generate two outputs using neural networks
To get N "O"utput vector dimensions of O corresponding to N "I"nput vector dimensions of I, the sizes of the input and target ma...

9 years ago | 0

| accepted

Answered
Custom Neural Netwrok (Manually re-implementing "patternnet' using "network")
H = 50 is excessive and probably causes overfitting/training problems. When starting something new, ALWAYS BEGIN with as man...

9 years ago | 1

Answered
High training error at the beginning of training the Convolutional neural network
>during the training process especially at the beginning of my training I >get extremely high training error after that this err...

9 years ago | 0

Answered
how do i create a multi-task learning neural network on matlab?
The number of output nodes is automatically determined by the number of rows in your target matrix. Similarly for for the input ...

9 years ago | 0

Answered
How to put a Matrix in the inputs of the neural network
Typically N I-dimensional "I"nput vectors are paired with N O-dimensional "O"utput target vectors and stored in matrices with si...

9 years ago | 0

| accepted

Answered
How to use a trained neural network in matlab??
If you have saved the net, try ynew = net( xnew ); enew = tnew - ynew; NMSEnew = mse( enew )/ mean( var( tnew',1 )...

9 years ago | 0

Answered
Our function is f(x): = x² Input Vector X = [ 0 1 2 3 4 5] Output Vector Y= [ 0 1 4 9 16 25] Prediction Vector Z= [ 10 9 8 7 6] The System Should predict the squared values of the Prediction Values. Thus, System have to give an a
Looks like homework. I did this problem when I first learned how to use the MATLAB NN Toolbox. However the input was [0:10]. ...

9 years ago | 0

Answered
Number of epoch equals number of loop iterations
It doesn't make any sense to record every weight after every epoch. The net will automatically carry the last updated weights...

9 years ago | 0

Answered
Where can I find the MATLAB code for Principle Component Analysis and MLP ANN to carry out Fault Diagnosis in rotating machines?
SEARCH both the NEWSGROUP and ANSWERS with the search words principal component analysis *Thank you for formally acceptin...

9 years ago | 0

Answered
Neural Netoworks Timestep Simulation
The net has to LEARN how to do the estimation. Therefore, you have to begin with known paired samples of input and corresponding...

9 years ago | 0

Answered
How to interpret Neural Network output if it is NaN?
For classification into mutually exclusive classes, the target columns should be [0,1] unit vectors. For examples, search the N...

9 years ago | 0

Answered
Which activation function is used by the Matlab Convolutional Neural Network Toolbox for the Fully-Connected-Layer?
The MATLAB default converts inputs and targets to the [ -1, 1] range. However, this does not help in detecting outliers that sho...

9 years ago | 0

| accepted

Answered
NARX Neural Network Tool not actually predicting?
You forgot to initialize the net with nonzero delay values. The obvious choice is the last two values of the val subset: 1. L...

9 years ago | 1

| accepted

Answered
Problem with bias-variance tradeoff in neural network
I don't see how a net trained on a discontinuous target could be very useful for studying the error bias/variance tradeoff. ...

9 years ago | 0

| accepted

Answered
How to create Multiple Output Neural Networks
It is very simple: For N I-dimensional "I"nputs paired with N O-dimensional "O"utput targets [ I N ] = size(inputmatrix) ...

9 years ago | 0

| accepted

Answered
How to predict 1 day ahead using NNTOOL
Insufficient information. I assume NARNET Which data set? N = ? Default delays 1:d, d = 2 ? input = x( 1 : N -...

9 years ago | 1

Answered
Problem with bias-variance tradeoff in neural network
Insufficient explanation: I assume classification. For which the mean and variance of the error e = t-y is typically, irrele...

9 years ago | 0