Klasifikasi Ras Kucing Menggunakan Algoritma Convolutional Neural Network

KOwners' lack of knowledge about pet cat breeds can lead to errors in providing appropriate care. This study aims to develop a cat breed classification system using a Convolutional Neural Network (CNN) algorithm to recognize ten types of cat breeds: Angora, Bengal, British Shorthair, Domestic, Maine Coon, noncat, Persian, Ragdoll, Siamese, and Sphynx. The image dataset was obtained from the Kaggle website, then through the stages of preprocessing, data augmentation, training, and model evaluation, the system was built using the Xception architecture with a transfer learning approach. The trained model was integrated into an Android-based mobile application supported by the Flask backend. The evaluation results showed that the developed CNN model was able to achieve an accuracy of 90.87%. This application is not only able to classify cat breeds quickly but also provides information regarding the care and needs of the detected breeds. The results of this study are expected to provide practical and educational solutions for cat owners in recognizing and caring for their pets more precisely.