Penerapan Metode CNN (Convolutional Neural Network) Dalam Mengklasifikasi Jenis Ubur-Ubur

Sandy Andika Maulana
Shabrina Husna Batubara
Tasya Ade Amelia
Yohanna Permata Putri Pasaribu

Jurnal Penelitian Rumpun Ilmu Teknik
Vol. 2 No. 4 (2023): November : Jurnal Penelitian Rumpun Ilmu Teknik

The purpose of this research is to apply the Convolutional Neural Network (CNN) method to classify various types of jellyfish. Jellyfish as sea creatures have a variety of shapes and sizes. This research includes data acquisition, data pre-processing, classification, and evaluation. The Keras Sequential model was chosen to implement the CNN model in this study. The results of the study showed an accuracy rate of 87%. In addition, the CNN model training accuracy rate reached 0.9037 with a loss value of 0.2097, while in CNN model testing, the accuracy rate reached 0.7944 with a loss of 0.5228.