try the rattle package in R. It will make it easier for you to implement SVM and ANN.

You may give a sight to the free Weka software.

I used neuro- solutions software , it is very useful and easy to understand and implement any type of problem.

https://www.researchgate.net/publication/276920535_Classification_of_Mental_Disorders_Figures_based_on_Soft_Computing_Methods

Article Classification of Mental Disorders Figures based on Soft Com...

Thanks all for help but I m using libsvm 3.20 and matlab r2010a 

Try to use Weka software (SMO from function classification part).

sir i m m.tech student and i have to use matlab

Try "Weka", "R" software .. 

or even more advanced programming language like "Matlab" 

The clasification  of data can be more eficiently by using  vector machine and testing by corresponding neural network algorithm.

pero radonja.....u r right ..but how could i do it using libsvm in matlab.

I did not use "libsvm"  from matlab!

sir , i have submitted my synopsis report to university

now i couldn't change tools..so i have to use libsvm and matlab

If you want to understand svm , then i suggest you to spend some time on watching vedio lectures of machine learning by andrew g at coursera.org .https://www.coursera.org/course/ml

We can use SVM-light: 

http://svmlight.joachims.org/

I think the easy way to use the Weka software .SMO  for training a support vector classifier.