its easy:

Class = knnclassify(data_test,data_train,label_train,k);

C=confusionmat(label_test,Class);

acc_of_knn=sum(diag(C))/sum(C(:));

Dear Hamed

to test the performance of a classifier, you should provide some data which their class are known. The number of data which  truly classified by that classifier to the whole number of data will show you the accuracy. 

A good overview of the knn classifier in relation to other clasifiers is given in

S. Nedevschi, Strateties for dealing with real world clasification problems, Ph.D. thesis:

http://users.utcluj.ro/~cameliav/documents/TezaFinalLemnaru.pdf

Dr. Nedevschi obverves:

The kNN classifier does not show remarkable results in any of the cases studied, which makes them suitable only for baseline problem assessment (p. 68).

% Classify the fisheriris data with a K-Nearest Neighbor classifier

load fisheriris

c = knnclassify(meas,meas,species,4,'euclidean','Consensus');

cp = classperf(species,c)

get(cp)

% 10-fold cross-validation on the fisheriris data using linear

% discriminant analysis and the third column as only feature for

% classification

load fisheriris

indices = crossvalind('Kfold',species,10);

cp = classperf(species); % initializes the CP object

for i = 1:10

test = (indices == i); train = ~test;

class = classify(meas(test,3),meas(train,3),species(train));

% updates the CP object with the current classification results

classperf(cp,class,test)

end

cp.CorrectRate % queries for the correct classification rate

its easy:

Class = knnclassify(data_test,data_train,label_train,k);

C=confusionmat(label_test,Class);

acc_of_knn=sum(diag(C))/sum(C(:));