The new type of Coronavirus (COVID-19), which started in 2019 in Wuhan, China, is an infectious virus that causes respiratory tract infection. This virus became effective in the world in a short time and turned into an epidemic. Early diagnosis of such infectious diseases and initiation of the necessary treatment at an early stage are very important. The use of X-ray (X-Ray) and Computed Tomography (CT) medical radiological imaging methods and deep learning and machine learning techniques help in the accurate and rapid detection of this disease. In this study; Two different datasets were used, including X-Ray images labeled normal-COVID-19-pneumonia (pneumonia) and CT images labeled normal-COVID-19. Inception ResNetV2, VGG-16 and DenseNet121 deep learning architectures and kNN and SVM classifiers are used. In this context, 3 different experiments were carried out. First of all, the classification performance of each network was examined. Then, the feature vectors produced by the networks were separately processed with classifiers. Finally, the feature vectors produced by the networks were combined and the classification process was carried out. As a result, the highest result for COVID-19 and normal images in the chest CT dataset was obtained with the combined features and kNN classifier with 98.9% accuracy.