How can i change the kernel in SVM classification

By using linear kernel I got the result using svmtrain() and svmclassify() function. But the result obtained is not so accurate. When I used rbf_kernel and polynomial I got an error as follows. Please help me how to use parameters with example...

svmStruct = svmtrain(X(P.training,:),Y(P.training),'showplot',true);

(I know the svmStruct could be change with :

SVMStruct = svmtrain(X(P.test,:),Y(P.test),'Kernel_Function', 'polynomial', 'Polyorder', 3);

C = svmclassify(svmStruct,X(P.training,:),'showplot',true);

but, what about the C??? Because i've tried so many times but still error :(

Can you guys help me? thanks