Yılmaz, A.A.Sevinç, Ö.2025-03-052025-03-0520242149-914410.22399/ijcesen.9632-s2.0-85218076213https://doi.org/10.22399/ijcesen.963https://hdl.handle.net/20.500.14411/10482The infectious coronavirus disease (COVID-19), seen in Wuhan city of China in December 2019, led to a global pandemic, resulting in countless deaths. The healthcare sector has become extensively use of deep learning (DL), a method that is currently quite popular. The aim of this study is to identify the best and most successful deep learning model and optimizer approach combination for COVID-19 diagnosis. For this reason, several DL methods and optimizer techniques are tested on two comprehensive public data set to select the best DL model with optimizer technique. A variety of performance evaluation metrics, including f-score, precision, specificity, and accuracy, were used to assess the models' effectiveness. The experimental results show that the most suitable and effective architecture is DenseNet-201 in the network comparison, which achieved a 98% accuracy rate using the AdaGrad optimizer and 200 iterations. © IJCESEN.eninfo:eu-repo/semantics/openAccessCovid-19Ct ImagesDeep LearningDeep Neural NetworkArticleN/AQ4111816826