Implementasi Metode Convolutional Neural Network (CNN) untuk Klasifikasi Jenis Ras Kucing

Aliefah Syalma Ratsdea Muftti (Universitas Muhammadiyah Ponorogo)
Yovi Litanianda (Universitas Muhammadiyah Ponorogo)

Uranus: Jurnal Ilmiah Teknik Elektro, Sains dan Informatika
Vol. 2 No. 2 (2024): Juni: Jurnal Ilmiah Teknik Elektro, Sains dan Informatika

This research implements the Convolutional Neural Network (CNN) method to classify the various types of cat breeds that are common in Indonesia. This research attempts to create an automatic system that can definitely and accurately classify and identify the types of cat breeds that exist in Indonesia using image processing techniques. The data used contains a total of 600 images with each folder containing 200 images. Using this CNN method produces a validation accuracy rate of 54% in the process of classifying cat breeds. Research shows that further developing the image processing process will increase the accuracy value of the resulting system.