Claim Missing Document
Check
Articles

Found 3 Documents
Search

ENSEMBLE CNN WITH ADASYN FOR MULTICLASS CLASSIFICATION ON CABBAGE PESTS Sovia, Nabila Ayunda; Wardhani, Ni Wayan Surya
BAREKENG: Jurnal Ilmu Matematika dan Terapan Vol 18 No 2 (2024): BAREKENG: Journal of Mathematics and Its Application
Publisher : PATTIMURA UNIVERSITY

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30598/barekengvol18iss2pp1237-1248

Abstract

Image classification is a complex process influenced by various factors, one of which is the amount of image data. In the context of cabbage pest classification, data often exhibits a significant class imbalance, where certain pests are more prevalent than others. This imbalance can pose challenges during model training and evaluation, potentially leading to biases in favor of the majority pests and reduced accuracy in identifying and classifying the less common ones. This research aims to enhance the classification performance for multiclass data specific to cabbage pests. We propose an ensemble learning approach that combines Convolutional Neural Network (CNN), Support Vector Machine (SVM), and Bagging methods. To address the imbalance issue inherent in cabbage pest data, we employ the Adaptive Synthetic Sampling (ADASYN) resampling technique. The CNN acts as the primary image identifier and classifier for various cabbage pests. Subsequently, the CNN model is integrated into SVM and Bagging models to mitigate the challenges of imbalanced data in pest classification. The research outcomes demonstrate that our ensemble approach, in conjunction with the ADASYN resampling technique, achieves an impressive accuracy rate of 97%, signifying its potential for improved cabbage pest detection and classification.
Model Hibrida CNN Berbobot dan Model Adaboost Decision Tree untuk Klasifikasi Penyakit Kubis pada Dataset Tidak Seimbang Iriany , Atiek; Sovia, Nabila Ayunda; Wardhani, Ni Wayan Surya; Sumarminingsih, Eni
Jurnal Teknologi Informasi dan Ilmu Komputer Vol 12 No 5: Oktober 2025
Publisher : Fakultas Ilmu Komputer, Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25126/jtiik.2025125

Abstract

Metode klasifikasi berbasis gambar banyak digunakan dalam bidang pertanian untuk mendeteksi penyakit pada berbagai tanaman, termasuk jenis yang rentan terhadap infeksi seperti kubis. Namun, performa optimal dalam klasifikasi gambar sangat dipengaruhi oleh jumlah dan keseimbangan data. Ketidakseimbangan data dalam klasifikasi penyakit tanaman kubis dapat menyebabkan model lebih memfokuskan perhatian pada kelas mayoritas, sehingga mengabaikan kelas minoritas, terutama pada klasifikasi multi-kelas. CNN sering digunakan dalam klasifikasi gambar, tetapi memiliki kelemahan dalam menangani data tidak seimbang karena cenderung lebih fokus pada kelas mayoritas. Untuk mengatasi permasalahan ini, dikembangkan model Hybrid yang mengombinasikan metode pembobotan pada CNN untuk ekstraksi fitur, model estimasi menggunakan Decision Tree, serta teknik ensemble Adaboost pada tahap klasifikasi. Pendekatan ini dirancang untuk meningkatkan kemampuan generalisasi terhadap kelas minoritas serta menghasilkan distribusi prediksi yang lebih seimbang. Hasil penelitian menunjukkan bahwa model Hybrid yang diusulkan mampu meningkatkan performa klasifikasi sebagaimana terlihat dari peningkatan Weighted Average F1-score menjadi 97%, dibandingkan model CNN tunggal dengan pembobotan yang hanya mencapai 63%. Peningkatan ini menunjukkan bahwa model Hybrid tidak hanya lebih efektif dalam menangani ketidakseimbangan data, tetapi juga mampu melakukan generalisasi yang lebih baik.   Abstract Image-based classification methods are widely used in agriculture to detect diseases in various plants, including those susceptible to infections, such as cabbage. However, achieving optimal performance in image classification is highly influenced by the quantity and balance of the data. Data imbalance in cabbage disease classification can cause the model to focus more on the majority class while neglecting the minority class, especially in multi-class classification. CNN is commonly used for image classification but struggles with imbalanced data, as it tends to prioritize the majority class. To address this issue, a hybrid model has been developed by combining weighting techniques in CNN for feature extraction, a Decision Tree for estimation, and the Adaboost ensemble technique for classification. This approach is designed to enhance generalization for minority classes and produce a more balanced prediction distribution. The results of the study indicate that the proposed Hybrid model is capable of improving classification performance, as evidenced by an increase in the Weighted Average F1-score to 97%, compared to the weighted CNN model, which achieved only 63%. This improvement demonstrates that the Hybrid model is not only more effective in addressing data imbalance but also better at generalizing the data.
Enhancing Image Classification of Cabbage Plant Diseases Using a Hybrid Model Convolutional Neural Network and XGBoost Sovia, Nabila Ayunda; Wardhani, Ni Wayan Surya; Sumarminingsih, Eni; Shofa, Elvo Ramadhan
CAUCHY: Jurnal Matematika Murni dan Aplikasi Vol 10, No 1 (2025): CAUCHY: JURNAL MATEMATIKA MURNI DAN APLIKASI
Publisher : Mathematics Department, Universitas Islam Negeri Maulana Malik Ibrahim Malang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.18860/cauchy.v10i1.30866

Abstract

Classifying imbalanced datasets presents significant challenges, often leading to biased model performance, particularly in multiclass classification. This study addresses these issues by integrating Convolutional Neural Networks (CNN) and XGBoost, leveraging CNN’s exceptional feature extraction capabilities and XGBoost's robust handling of imbalanced data. The Hybrid CNN-XGBoost model was applied to classify cabbage plants affected by pests and diseases, which are categorized into five classes, with a significant imbalance between healthy and affected plants. The dataset, characterized by severe class imbalance, was effectively handled by the proposed model. A comparative analysis demonstrated that the CNN-XGBoost approach, with a Balanced Accuracy of 0.93 compared to 0.53 for the standalone CNN, significantly outperformed the standalone model, particularly for minority class predictions. This approach not only enhances the accuracy of plant disease and pest diagnosis but also provides a practical solution for farmers to efficiently identify and classify cabbage plants, contributing to more effective agricultural management.