Enhancing the identification of date palm diseases with ensemble learning techniques
Abstract
Date palms (Phoenix dactylifera L.) are one of the most important crops in the arid and semi-arid regions, providing food, shelter, and economic benefits. However, plant disease identification is an important challenge in the agriculture sector and it is critical to mitigating crop loss and disease spread. Early detection of palm diseases is at the core of our study. In this work, we identify and classify three nutrient deficiencies and five diseases using a set of convolutional neural network (CNN) architectures (VGG16, VGG19, MobileNet, MobileNetV2, DenseNet169, and Xception). The proposed approach is divided into two main phases. In the first phase, we apply transfer learning and a fine-tuning process, which allows the networks to adjust to the specific features. During the second phase, two ensemble learning techniques are explored: voting and bagging, each helping to increase generalization and reduce prediction variance. In bagging, we apply the feature fusion of fine-tuned models, which enables us to achieve an accuracy of 99.44% using the decision tree classifier. These results surpass recent and existing works in this field, demonstrating the effectiveness of ensemble learning in leveraging model complementarity to enhance robustness and accuracy in identifying palm diseases.
Keywords
Bagging; Convolutional neural network; Ensemble learning; Fine-tuning; Palm diseases; Voting
Full Text:
PDFDOI: https://doi.org/10.11591/eei.v15i6.12128
Refbacks
- There are currently no refbacks.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Bulletin of Electrical Engineering and Informatics (BEEI)
ISSN: 2089-3191
,
e-ISSN: 2302-9285
This journal is published by the
Institute of Advanced Engineering and Science (IAES)
.