Answered
ERROR: A regression layer must not be preceded by a softmax layer.
Softmax is designed to be a classifier output component that yields a classification probability output with components in [0 1 ...

7 years ago | 0

Answered
Problem with mathematical expression of neural network
You can answer that question better than we can. Normalize your data and see what happens. Or am I missing something ??? ...

7 years ago | 0

Answered
How to build a neural network which is not Fully-connected with NN toolbox?
The best approach is to find, via an exhaustive search within bounds, the minimum number of hidden nodes that will yield your de...

7 years ago | 0

Answered
How can I use Neural network to classify signals of four objects?
I assume that you are not referring to time series. For classification of c classes from N measurements of I-dimensional vect...

7 years ago | 0

| accepted

Answered
Neural network training function error modulo a number
The simplest function that makes perfect sense is the corresponding squared Euclidean distance between corresponding points on t...

7 years ago | 0

Answered
How can i forecast future production using trained time series data?
1. Format your code so that it will run when copied and pasted. 2. Test your code on the documentation example. 3. If step ...

7 years ago | 0

Answered
my input data size is converting automatically in different data size.
You made at least one mistake. However, since you didn't include your code, you will have to wait until the mind readers come...

7 years ago | 0

Answered
Where can i find the algorithm of ANN toolbox in MATLAB?
The NN toolbox contains many algorithms. Each algorithm has HELP and DOC documentation with additional references. A good wa...

7 years ago | 0

Answered
How clustering analysis improve neural networks performance ?
Clustering is pivotal in designing more accurate elliptical basis function classifiers. In particular, this allows centering,...

7 years ago | 0

| accepted

Answered
Can any one provide step by step procedure for trainlm and trainbr algorithm
1. Read the documentation a. help trainbr b. doc trainbr c. type trainbr 2. Consult posts in comp.soft-sys.matlab...

8 years ago | 0

Answered
MSE result is too high
1. GUESS: fitnet ? 2. Unnecessary assignments of default values to divideFcn and ratios. 3. Erroneous use of "output...

8 years ago | 0

Answered
I need a program of neural network mlp containing 4 entries and one output
Use the help and doc commands for basic examples 1. Regression/Curvefitting? help FITNET and doc fitnet ...

8 years ago | 0

Answered
When training a NARX network during the validation step, does the network use the outputs it calculates to feedback ? Or does it use the outputs you provide ?
It uses both: MODE Open-Loop (OL) Starts with TARGETS in the feedback connections Close-Loop (CL) Finishes with OUTPUT...

8 years ago | 0

| accepted

Answered
Neural Network - How to use a custom training function
Use the help and doc commands on trainru trainb trainr trainc trains trainbu trainscg traingdx traingdm traingd tr...

8 years ago | 0

Answered
How to best present multi-parameter, multi-read, multi-sample data for neural net learning?
Both target and inputs should be as mixed as possible. Something close to S1,S3,S2,S3,S1,S3,S2,S3,... should work well. ...

8 years ago | 0

Answered
Brief explanation of all the training function in nntool
If you have a regression or curve-fitting problem use training function FITNET. Accept all defaults and start with the code fro...

8 years ago | 0

Answered
Alpha-numeric or alphabetic data type format for Neural Network Toolbox (nntool)
alpha_numeric Hope this helps *Thank you for formally accepting my answer* Greg

8 years ago | 0

Answered
What is the best way to take advantage of hints or preexisting knowledge between classes and features when training a patternnet?
Too much overthinking. Just train as usual, the net will be smart enough to do the sorting. When I say train as usual I me...

8 years ago | 0

Answered
comment faire la classification
Start with the documentation help patternnet doc patternnet Then followup with examples in both the NEWSGROUP i...

8 years ago | 0

| accepted

Answered
How to plot a confusion matrix for multi-class categorical labels?
Convert the categorical labels to a binary equivalent.

8 years ago | 0

Answered
suddenly and together train by MATLAB
I have had very few problems with a double for-loop approach that relies on default parameters except for 1. OUTER-LOOP o...

8 years ago | 0

Answered
Neural Networks Toolbox: input layer with zerocenter normalization, where to find values of normalizing matrix?
All net properties can be deduced by omitting the ending semicolon in the command net = net *Thank you for formally accep...

8 years ago | 0

Answered
Can I use parameters of different network training functions with each other?
My approach for successfully designing hundreds of nets is straightforward: Accept all defaults except 1. Initial state o...

8 years ago | 0

Answered
RBF network not generalizing well
Ntrn ~ 0.7*2500 = 1750 samples span, AT MOST, a 1,749 dimensional input space. Therefore, an input dimension of 40,000 is rid...

8 years ago | 1

Answered
What is the default transfer function for fitnet?
%If you want to know network details, just remove ending semicolons on a succession of commands net = fitnet %Then scroll...

8 years ago | 0

Answered
In MATLAB 2017b, the patternnet example for classification of iris data no longer works it seems. I've tried on multiple machines. But it does work on MATLAB 2016b. What is going on with this? And how do I actually get it to work.
Mine works: close all, clc rng(0), version % '9.3.0.713579 (R2017b)' [x,t] = iris_dataset; net = patternnet(10);...

8 years ago | 1

Answered
Neural Network Transfer Function
There is no good reason to differ from the default help fitnet % Form of feedforwardnet for regression & curvefitting) ...

8 years ago | 1

Answered
Predict one independent variable (IV) from 25 dependent variables (DV)
Even if you do not want to use a net, I doubt if you need all 25 to model 1. I have successfully used the following strategy:...

8 years ago | 0

| accepted

Answered
is it possible to touch the weights & bias in neural network toolbox
% OMIT THE ENDING SEMICOLONS IN THE FOLLOWING 2 COMMANDS net = fitnet TRAINPARAM = net.trainParam % You can directly mo...

8 years ago | 0

Answered
electricity price forecasting using neural network
Start with the default code. Run 10 times with different initial weights. If successful, repeat with smaller number of hidden...

8 years ago | 0

| accepted