Answered
How I can calculate prediction interval for neural network
Assume errors are Gaussian distributed with zero mean and std = sqrt(mse). Hope this helps. *Thank you for formally accept...

12 years ago | 0

| accepted

Answered
net reusing
With the current set of net creation functions ( e.g., fitnet, patternnet, feedforwardnet,...) weight initialization does not o...

12 years ago | 1

Answered
what should be correct format of feature vectors matrix for feeding into neural networks?
For classification of c classes use N pairs of I-dimensional column vector inputs and O-dimensional outputs with O = c.. The out...

12 years ago | 3

| accepted

Answered
How many times should I re-train my images in nprtool ?
Data = Designdata + Testdata Designdata = Trainingdata + Validationdata % Performance estimates are BIASED If the data set...

12 years ago | 0

| accepted

Answered
How much number of hidden neurons should I take in nprtool ?
It is always best to 1. Accept all defaults 2. Run 3. Evaluate the non-training validation and test results 4. If sat...

12 years ago | 1

| accepted

Answered
how to input different variables as input in Neural Network
Use an input matrix, x, with N 4-dimensiomal column vectors. Use a target matrix, t, with a 2-dimensional matrix of ones and ze...

12 years ago | 0

| accepted

Answered
How to code up Neural Networks ?
Once you have the input matrix of N I-dimensional column vectors and target matrix of N c-dimensional column unit vectors you ca...

12 years ago | 0

Answered
HOW TO VIEW NO OF HIDDEN NEURONS IN A SAVED NETWORK?
1. The net diagram view(net) 2. I = size(input,1), O = size(target,1), Nw = net.numWeightElements Assuming no input o...

12 years ago | 0

| accepted

Answered
Gradient Descent Neural Network
1. Image is not attached 2. Type net = net % w/o ending semicolon and find net.performanceFcn = 'mse' % i.e., the averag...

12 years ago | 0

| accepted

Answered
i hv attached the script generated for 2 layer(1 hidden layer) NN , what changes do i need to make to use it for NN with more than 1 hidden layer.
hiddenLayerSize = [10 5 ] net = patternnet(hiddenLayerSize); Hope this helps. *Thank you for formally accepting my ans...

12 years ago | 1

| accepted

Answered
How could I create a customised performance function for a neural network ?
You have the misleading impression that unbalanced data requires changing the minimization objective function. It does not. ...

12 years ago | 0

Answered
Help in viewing the output of a neural network
> Help in viewing the output of a neural network > Asked by sundar on 2 Feb 2014 at 11:01 > I used nftool in Matlab 2012 ...

12 years ago | 1

| accepted

Answered
Neural Network - How to use a custom performance function
This may help: http://www.mathworks.com/matlabcentral/answers/56137-how-to-use-a-custom-transfer-function-in-neural-net-train...

12 years ago | 2

| accepted

Answered
Neural Network - How to use a custom performance function
Check this out http://www.mathworks.com/matlabcentral/answers/56137-how-to-use-a-custom-transfer-function-in-neural-net-train...

12 years ago | 1

Answered
What is the activation function of Levenberg–Marquardt Backpropagation algorithm?
I don't think there is any limit on which differentiable performance functions are used except the output function must allow th...

12 years ago | 1

| accepted

Answered
How could I create a customised performance function for a neural network ?
This may be of interest: http://www.mathworks.com/matlabcentral/answers/56137-how-to-use-a-custom-transfer-function-in-neura...

12 years ago | 0

Answered
How to code up Neural Networks ?
I suggest using a standard classifier with patternnet, one hidden layer, and target columns from eye(23). Use as many defaul...

12 years ago | 0

| accepted

Answered
i have excel of form 7*12, i want to use this data in neural network
net = narxnet(ID,FD,H); net.divisionFcn = 'divideblock' Hope this helps. *Thank you for formally accepting ...

12 years ago | 1

| accepted

Answered
sir/ madem i want to forecast the weekly rainfall using ANN technique. please give me some demo on this.
help timedelaynet doc timedelaynet help narxnet doc narxnet Also, search the NEWSGROUP and ANSWERS using timede...

12 years ago | 0

| accepted

Answered
cross validation for neural network
Search the NEWSGROUP and ANSWERS using greg crossvalidation and greg cross-validation and greg cross valida...

12 years ago | 0

| accepted

Answered
How would I access the weights of a trained neural network?
1. Before the double loop initialize your RNG (e.g., rng(0) ). From this you can deduce the state of the RNG before loop (nopt,H...

12 years ago | 0

| accepted

Answered
How to train the neural network using RBF?
1. When you create newrb it automatically trains itself without the trn/val/set data division. Therefore, only include the train...

12 years ago | 2

Answered
What is the difference between 'divideblock' and 'divideint'?
Have you tried using the help and doc commands? Have you tried using simple examples without the ending semicolon? help d...

12 years ago | 1

| accepted

Answered
How to code up Neural Networks ?
I am confused by your question "Is there a need of nprtool to do that? Please clarify that statement. Comment The f...

12 years ago | 0

Answered
How to change Neural Network performance from MSE to MAE?
clear all; format compact load ShotData; % Contains A and B. %What are size(A) and size(B)? [rA cA] = size(A) [rB cB ] ...

12 years ago | 0

| accepted

Answered
Can somebody explain me how to use "divideind"??
{close all, clear all, clc [ x, t ] = simpleclass_dataset; [ I N ] = size(x) % [ 2 1000] [ O N ] = size(t)...

12 years ago | 1

Answered
Can somebody explain me how to use "divideind"??
Your original problem of nonrepeatibility is easily solved by initializing the RNG before it is used to divide data or initial...

12 years ago | 1

Answered
Prediction of future values using narnet
% The code: ==>When posting should reformat to have one code statement per line if true ==> What does that stateme...

12 years ago | 1

| accepted

Answered
How can I handle larger powers of 10 like 10^800 in MATLAB?
Use normalization and/or logarithms. *Thank you for formally accepting my answer* Greg

12 years ago | 0

Answered
Time Series Ahead Prediction in Neural Network, Large Scale Iterative Training
1. You can only predict as far as the target has significant autocorrelation lags and/or the input and target have significant c...

12 years ago | 0

| accepted