Dear Rajeshwari,

Hi !,

MATLAB is the powerful tool for research community.

If you are using "Neural network Tool" in MATLAB then I prefer to use following link: "100 Best MATLAB Neural Network Videos".

These videos are good for beginners.

http://meta-guide.com/videography/100-best-matlab-neural-network-videos

Thanks.

Rajesh Mahadeva

hi, Go to the following site

www.mathworks.com/help/nnet/ref/train.html 

hi

you can use 

>>nprtool

good luck 

Can't you use the neural network toolbox(nntool)? It is a gui, which allows you to import your data and export your outputs. 

Dear Rajeshwari B S,

You can use for work with neural networks in MATLAB from following functions :

train.m

newff.m

newcf.m

newelm.m

sim.m

init.m

adapt.m

Best regards

Alireza

If you prefers a "good book or a web site" to try implement by yourself using gradient descent, I would suggest the following MOOCs

http://openclassroom.stanford.edu/MainFolder/CoursePage.php?course=ufldl

sez. III. NEURAL NETWORKS

https://www.coursera.org/course/ml

VIII. Neural Networks: Representation (Week 4)

IX. Neural Networks: Learning (Week 5)

Hello,

I would recommend you start with 

Demuth H, Beale M. Neural network toolbox users guide. Math-

Works, Inc.; 2003.

It is very user friendly. Then go to the online Matlab's help for the NN ToolBOX.

Good luck!

Luiza

Hi,

If you are using nntool in matlabl, things are pretty easy. You can check its documentation for how to do. Basically, it will ask you to give a set of inputs and a corresponding set of outputs. All these, inputs and outputs, are the training datasets. MATLAB, would set aside a portion of these data for training and the rest for cross validation. Once the training is done, you'd be given the performance curves as well as the weights of the nn. You can save this network, and try to use it for processing other inputs.

Another approach, if you want to do it from scratch, is gradient descent optimization or other similar methods. A nice tutorial on using the gradient descent and how to check the convergence is explained amazingly by Prof. Andrew in Coursera Machine Learning course.

you can begin by using "nftool", but you should have data corresponding the inputs and the outputs of the neural network that you designed.

on other hand ,you can make use of this function :

net = newff(trainSet,t,n,{'tansig','logsig'},'trainscg');

tansig: activation function of hiden neurons

logsig:activation function of output neurons

trainscg: is the algorithm of training of the network, for more options make a look in the help of matlab.

but still, the help of matlab is the most useful tool to learn how to perform neural network in the said software.

Hello all

i am not the author for the paper " "How do I train a neural network in Matlab?"

For Matlab Code, Visit link

http://dg-algorithm.blogspot.com/

Hi, you can use nftool in MATLAB

Dear Rajeshwari,

Hi !,

MATLAB is the powerful tool for research community.

If you are using "Neural network Tool" in MATLAB then I prefer to use following link: "100 Best MATLAB Neural Network Videos".

These videos are good for beginners.

http://meta-guide.com/videography/100-best-matlab-neural-network-videos

Thanks.

Rajesh Mahadeva

Was looking the same question to be posted, thanks lot for the replies

Neural Network Control Of Robot Manipulators And Non-Linear Systems

please refer this book

https://books.google.co.in/books/about/Introduction_to_Neural_Networks_Using_Ma.html?id=jJTN8RPgyXgC

Dear Rajeshwari B S

Mathworks explain the type of training neural network .

https://www.mathworks.com/help/deeplearning/ug/neural-network-training-concepts.html;jsessionid=872e269c595b41519556e6f06a29

The MATLAB Neural Network Toolbox user guide is very helpful provided we've basic understanding of Neural Networks.

It is a tool for prediction applications wherever applicable. It will give a relationship between targeted output and all other input parameters and also how the parameters influence the target.