Pada penelitian ini menggunakan algoritma Convolutional Neural Network (CNN) untuk mendeteksi COVID-19 berdasarkan citra X-ray Paru-paru. Arsitektur CNN yang digunakan adalah EfficientNetB7 dan Resnet152V2 dengan memanfaatkan teknik Transfer Learning. Penelitian ini berfokus pada membandingkan kinerja kedua model arsitektur dalam mengklasifikasikan citra X-ray Paru-paru terinfeksi COVID-19. Selanjutnya mengimplementasikan model CNN tersebut ke aplikasi deteksi Citra X-ray paru-paru berbasis web. Dari hasil evaluasi kedua model tersebut disimpulkan bahwa Resnet152-V2 mencapai kinerja lebih baik dibanding arsitektur CNN EfficientNetB7 dengan akurasi 97% sedangkan EfficientNetB7 dengan akurasi 95%.

Jamu is one of Indonesia's cultural heritage, which consists of several plants that have been practiced for centuries in Indonesian society to maintain health and treat diseases. One of the scientification efforts of Jamu to reveal its mechanism is to predict the target-protein of the active ingredients of the Jamu. In this study, the prediction of the target compound for Jamu was carried out using a supervised learning approach involving conventional medicinal compounds as training data. The method used in this study is the closest profile method adopted from the nearest neighbor algorithm. This method is implemented in drug compound data to construct a learning model. The AUC value for measuring performance of the three implemented models is 0.62 for the fixed compound model, 0.78 for the fixed target model, and 0.83 for the mixed model. The fixed compound model is then used to construct a prediction model on the herbal medicine data with an optimal threshold value of 0.91. The model produced 10 potential compounds in the herbal formula and its 44 unique protein targets. Even though it has many limitations in obtaining a good performance, the closest profile method can be used to predict the target of the herbal compound whose target is not yet known.

The problem of unbalanced data is important in the field of Data Mining. Dataset with unbalanced classes is a dataset whose frequency of occurrence of certain classes is very much different from other classes. This imbalance problem will bias the classifier's performance. Many researchers have examined both the development of algorithms and modifications to the preprocessing stage to overcome this problem. This study discusses the comparison of One Class Classification algorithms, namely Elliptic Envelope and Isolation Forest on unbalanced data. From this study, the Elliptic Envelope Method showed better results compared to the Isolation Forest method with 80.28% recall testing and 80.28% precision while Isolation Forest showed 46.95% recall results and 46.95% precision.