Deep Learning based Prediction of Crop Disease in Federated Learning Environment
Journal of Intelligent Decision Making and Information Science · 27.08.2026 · 3 (9s) · 1354-1376
DOI: 10.17762/jidmis.v3.2488 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Intelligent Decision Making and Information Science, online erschienen am 27.08.2026. Verfasst von Sadananda Behera. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Manual monitoring and detecting diseases in plants are nearly impossible. It needs a lot of time and a skilled, knowledgeable person. This study proposes a deep learning-based crop disease prediction framework within a Federated Learning (FL) environment. In this framework, multiple clients collaboratively train the models without sharing raw data. The study utilizes the publicly available PlantVillage dataset, which contains labelled images of healthy and diseased crop leaves. The dataset is further divided into training and testing subsets to evaluate different deep learning models within the federated learning environment. The proposed framework also evaluates federated learning aggregation algorithm, such as FedAvg, in combination with different deep learning models. These FL algorithms are used to check the effectiveness of privacy-preservation in crop disease prediction. This study is performed on eight Deep Learning (DL) models such as CNN, RNN, LSTM, GRU, EfficientNet, Transformer, MobileNet, and ResNet in standalone and FL environment. Key qualities include decentralized computation, data privacy, and data sovereignty, as raw datasets from clients are never transferred to a central server. It has been observed that the CNN model achieved an accuracy of 99%, precision of 94%, recall of 95%, F1-score of 93% and AUC-ROC score of 97% for crop disease prediction under Standalone mode while scored 95% in privacy preservation. In other hand, EfficientNet though scored 97% accuracy, scored 97% in privacy preservation. There is slight performance drop (2 – 5%) in FL settings for all the models. Considering the results EfficientNet in FL setting is the best choice to consider for digital data privacy and data sovereignty prospective.
Zitieren
APA
Behera, S. (2026). Deep Learning based Prediction of Crop Disease in Federated Learning Environment. Journal of Intelligent Decision Making and Information Science, 3(9s), 1354-1376. https://doi.org/10.17762/jidmis.v3.2488
Vancouver
Behera S. Deep Learning based Prediction of Crop Disease in Federated Learning Environment. Journal of Intelligent Decision Making and Information Science. 2026;3(9s):1354-1376. doi:10.17762/jidmis.v3.2488
BibTeX
@article{Behera2026,
author = {Behera, Sadananda},
title = {Deep Learning based Prediction of Crop Disease in Federated Learning Environment},
journal = {Journal of Intelligent Decision Making and Information Science},
year = {2026},
volume = {3},
number = {9s},
pages = {1354--1376},
doi = {10.17762/jidmis.v3.2488}
}
EUVIMED