DETEKSI COVID-19 PADA CITRA X-RAY DADA MANUSIA MENGGUNAKAN TRANSFER LEARNING VGG19

Coronavirus Disease (COVID-19) is a new disease originating in the city of Wuhan, China, which emerged at the end of 2019 until now. This disease has hit almost the whole world by infecting the respiratory system, ranging from the common cold to serious diseases, such as Middle East Respiratory Syndrome (MERS) and Severe Acute Respiratory Syndrome (SARS). In Indonesia, several methods have been carried out to detect COVID-19, such as the Polymerase Chain Reaction (PCR) and Rapid tests. The method used to detect COVID-19 currently takes a long time to get accurate results in diagnosis. This study aims to classify human chest X-ray images using the Transfer Learning method with the VGG19 model. The applied model is also accompanied by data augmentation and data balancing techniques using undersampling techniques. The dataset used is the "COVID-19 Radiographic Database (COVID-19 and Normal)" with a total of 13,808 data divided into 2 classes, namely, COVID-19 and Normal. The built model has accuracy, precision, recall, and confusion matrix values. The accuracy results generated by the VGG19 model are 90%