Article Per Year (5 Year)
p-Index From 2021 - 2026
0.23
P-Index
This Author published in this journals
All Journal JITK (Jurnal Ilmu Pengetahuan dan Komputer)
Attarik Mohammad
Ahmad Dahlan University
Computer Science & IT

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
PERFORMANCE EVALUATION OF TRANSFER LEARNING MODELS BASED ON OPTIMIZATION IN AGRICULTURAL PEST CLASSIFICATION Attarik Mohammad; Sugiyarto Surono; Aris Thobirin
JITK (Jurnal Ilmu Pengetahuan dan Teknologi Komputer) Vol. 11 No. 4 (2026): JITK Issue May 2026
Publisher : LPPM Nusa Mandiri

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33480/jitk.v11i4.7800

Abstract

Pests in agriculture lower crop yields and jeopardize the world’s food security. Thus, quick and precise pest identification is crucial for successful pest management. Convolutional Neural Networks (CNN) and other deep learning techniques have made it possible to automatically classify pests thanks to developments in digital image processing and artificial intelligence (AI). Using three optimization algorithms, Adam, RMSprop, and SGD, this study assesses three transfer learning architectures, ResNet50V2, Xception, and EfficientNetB0. This study’s primary contribution is a comparative analysis of CNN architectures and optimization techniques to determine the best configuration for classifying agricultural pests. The dataset, which includes 5494 pest photos from 12 classes, was acquired via Kaggle. A ratio of  80%, 10%, and 10% was used to separate the data into training, validation, and testing sets. The performance of feature extraction and classification was enhanced by applying transfer learning with fine-tuning. According to findings, Xception with Adam and RMSprop has the highest accuracy of 94%. Adam and EfficientNetB0 both achieved competitive results with the same precision. These results suggest that the performance of agricultural pest classification models is influenced by both optimizer and architecture choices.
Co-Authors Aris Thobirin Sugiyarto Surono