Dear Supriya

describe your needs, you want to know svm matlab  implementation? or ANN&SVM implementation?

i have libsvm matlab interface. i have done classification by libsvm in matlab and by ANN too.

but now i want to train prediction model by libsvm and convert that to equivalent ANN model and then do prediction by that ANN model.

I'm not understand your work exactly but i suggest for you my friend:

train your data by SVM,unsupervised and then train ANN supervised.

I hope this would help you.

I still don't think this question makes sense.  Can you explain a little about why you would want to do this?

thanks Mohammad sir,

exactly i want to do this only....but i dont know how could i do.

i want to train my data by SVM , and take the support vectors to input to ANN to train prediction model and thn want to predict my test dataset by this ANN model.

Dear Supriya,

when you are using SVM as a clustering method, it would return output, use that output to classification that in ANN, in your ANN input is same as SVM, and the output is what you get from SVM.

input -> SVM (result as output)

[input output] -> ANN -> classification and train it

but mention the check data too, use about 70% of your data for training SVM and ANN, then use 30% for checking the result in ANN and how it's would classifier.

best regards

Thanku sir, 

So nice of u☺

sir , how could i give svm output to input to ann??

i dont know how to change format of svm's output to nn input format.