HYPERPARAMETER OPTIMIZATION ON THE CONVOLUTIONAL NEURAL NETWORK FOR ROBUSTA COFFEE FRUIT DISEASE IDENTIFICATION USING ARTIFICIAL BEE COLONY ALGORITHM

FADILAH, FAIZ (2025) HYPERPARAMETER OPTIMIZATION ON THE CONVOLUTIONAL NEURAL NETWORK FOR ROBUSTA COFFEE FRUIT DISEASE IDENTIFICATION USING ARTIFICIAL BEE COLONY ALGORITHM. Tugas Akhir thesis, Informatics.

Text 5210411356_FAIZ FADILAH_Abstrak.pdf Download (125kB)

Abstract

Indonesia, as one of the world’s largest coffee producers, faces serious threats from extreme weather and crop diseases that can reduce production by up to 30% and damage 60% of robusta coffee yields. Early identification of fruit diseases is critical to minimizing losses and maintaining coffee quality. This study aims to enhance the performance of Convolutional Neural Networks (CNN) in classifying robusta coffee fruit diseases by optimizing hyperparameters using the Artificial Bee Colony (ABC) algorithm. The dataset used comprises 2,100 images categorized into three classes: Healthy Berry, Berry Borer, and Berry Damage. The research results demonstrate that ABC significantly improves CNN classification accuracy, achieving 97.14% accuracy with optimal hyperparameters: learning rate of 0.0008, three convolutional layers, 35 filters, kernel size of 4, 128 dense units, and a dropout rate of 0.474. In comparison, a baseline 24-layer CNN model only reached 18% accuracy. The optimized model consistently attained between 94.76% and 97.14% accuracy. These findings contribute to the development of metaheuristic-based optimization techniques in agricultural image classification and support efforts to improve coffee quality amid global climate challenges. Keywords: Artificial Bee Colony, Convolutional Neural Network, Hyperparameter, Optimization, Coffee Fruit Disease

Item Type:	Thesis (Skripsi, Tugas Akhir or Kerja Praktek) (Tugas Akhir)
Subjects:	T Technology > T Technology (General)
Divisions:	Fakultas Sains Dan Teknologi > S1 Informatika
Depositing User:	Kaprodi S1 Informatika UTY
Date Deposited:	16 Jul 2025 02:31
Last Modified:	16 Jul 2025 02:31
URI:	http://eprints.uty.ac.id/id/eprint/18158

Actions (login required)

View Item