parameter_combinations = grid_search(feature_selection_parameters,classifier_parameters)
for k = 1:number_of_parameter_combinations
    for i = 1:5
        data_test_cv = data_whole{i}
        data_train_cv = data_whole-data_test_cv
        feature_selected = Fscore(data_train_cv,feature_selection_parameters{k})  
        classifer_cv = classifier_construct(feature_selectedr,classifier_parameters{k})
        label_predict_cv{i} = predict(classifer_cv,data_test_cv)
    end
    acc_cv(k) = acc_calculate(label_actual_cv,label_predict_cv)
end
index_max_acc = max_index(acc_cv)
best_parameter = parameter_combinations{index_max_acc)
feature_final = Fscore(data_whole,best_feature_selection_parameter)
classifier_final = classifier_construct(feature_final,best_classifier_parameter)