Klasifikasi Penyakit Daun Padi Menggunakan Convolutional Neural Network dengan Arsitektur AlexNet

Niko Pirnando, Hendra Nata
Petrus, Johannes
Tinaliah, Tinaliah

MDP Student Conference
Vol 4 No 1 (2025): The 4th MDP Student Conference 2025

Rice is a major food crop crucial for global food security, but its growth faces challenges such as leaf diseases caused by bacteria and fungi. Manual classification of rice leaf diseases by farmers in the field has limitations in accuracy and efficiency. This study aims to classify rice leaf diseases to provide fast and accurate classification results for early treatment. A convolutional neural network (CNN) with the AlexNet architecture will be used with the rice leaf dataset. The dataset consists of 2000 images divided into four classes: Healthy, Brown Spot, Hispa, and Leaf Blast, with an 80:10:10 split ratio. Testing compares the Adam and SGD optimizers with batch sizes of 16, 32, and epochs of 10, 20, and 25. The highest classification accuracy achieved was 80% using the Adam optimizer weight, improving model performance.