Optimasi Hyperparameter EfficientNet-B1 untuk Klasifikasi Penyakit Bawang Merah Berbasis Citra Digital
DOI:
https://doi.org/10.51903/elkom.v19i1.3790Keywords:
EfficientNet, Klasifikasi Citra, Deep Learning, Penyakit Bawang Merah, Transfer LearningAbstract
Penyakit bawang merah merupakan salah satu faktor yang dapat menurunkan produktivitas dan kualitas hasil panen. Identifikasi penyakit secara manual sering kali memerlukan waktu dan bergantung pada keahlian pengamat. Penelitian ini bertujuan menganalisis performa arsitektur EfficientNet-B0 dan EfficientNet-B1 dalam klasifikasi penyakit bawang merah berbasis citra digital. Dataset yang digunakan berasal dari Onion Leaf Disease Dataset dan dibagi menjadi 2.700 data pelatihan, 332 data validasi, dan 201 data pengujian. Model dibangun menggunakan pendekatan transfer learning dengan bobot awal ImageNet serta diuji menggunakan variasi learning rate (0,0001 dan 0,001) dan dropout (0,3 dan 0,5). Hasil penelitian menunjukkan bahwa EfficientNet-B1 dengan learning rate 0,001 dan dropout 0,3 memberikan performa terbaik dengan training accuracy 99,94%, validation accuracy 89,13%, dan test accuracy 93,03% serta test loss 0,2572. Analisis confusion matrix menunjukkan bahwa model mampu mengklasifikasikan 187 dari 201 citra pengujian dengan benar. Hasil penelitian membuktikan bahwa EfficientNet-B1 efektif digunakan untuk klasifikasi penyakit bawang merah dan berpotensi mendukung sistem deteksi penyakit tanaman secara otomatis.
References
Bhujel, S., & Shakya, S. (2022). Rice leaf diseases classification using discriminative fine tuning and CLR on EfficientNet. Journal of Soft Computing Paradigm, 4(3), 172–187. https://doi.org/10.36548/jscp.2022.3.006
Liu, M., Liang, H., & Hou, M. (2022). Research on cassava disease classification using the multi-scale fusion model based on EfficientNet and attention mechanism. Frontiers in Plant Science, 13, 1088531. https://doi.org/10.3389/fpls.2022.1088531
Albahli, S., & Masood, M. (2022). Efficient attention-based CNN network (EANet) for multi-class maize crop disease classification. Frontiers in Plant Science, 13, 1003152. https://doi.org/10.3389/fpls.2022.1003152
Nazir, T., Iqbal, M. M., Jabbar, S., Hussain, A., & Albathan, M. (2023). EfficientPNet—An optimized and efficient deep learning approach for classifying disease of potato plant leaves. Agriculture, 13(4), 841. https://doi.org/10.3390/agriculture13040841
You, H., Lu, Y., & Tang, H. (2023). Plant disease classification and adversarial attack using SimAM-EfficientNet and GP-MI-FGSM. Sustainability, 15(2), 1233. https://doi.org/10.3390/su15021233
Ashmafee, M. H., Ahmed, T., Ahmed, S., Hasan, M. B., Jahan, M. N., & Rahman, A. B. M. A. (2023). An efficient transfer learning based approach for apple leaf disease classification. arXiv. https://arxiv.org/abs/2304.06520
Appdevop666. (2024). Onion leaf disease dataset. Hugging Face. https://huggingface.co/datasets/appdevop666/onion-leaf-disease
Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep learning. MIT Press.
Tan, M., & Le, Q. (2019). EfficientNet: Rethinking model scaling for convolutional neural networks. In Proceedings of the 36th International Conference on Machine Learning (ICML) (pp. 6105–6114). PMLR.
Hu, J., Shen, L., & Sun, G. (2018). Squeeze-and-excitation networks. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 7132–7141. https://doi.org/10.1109/CVPR.2018.00745
Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., & Chen, L. C. (2018). MobileNetV2: Inverted residuals and linear bottlenecks. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 4510–4520. https://doi.org/10.1109/CVPR.2018.00474
Tan, M., & Le, Q. V. (2019). EfficientNet: Rethinking model scaling for convolutional neural networks. Proceedings of the 36th International Conference on Machine Learning (ICML), 97, 6105–6114. https://proceedings.mlr.press/v97/tan19a.html
Published
Issue
Section
License
Copyright (c) 2026 Elkom: Jurnal Elektronika dan Komputer
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
