Belghit et al. (2026) Applying AdaBoost algorithm on multiclass OvA-SVM for the delineation of rainy clouds using multispectral MSG-SEVIRI data

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Springer Link (Chiba Institute of Technology)
• Year: 2026
• Date: 2026-03-20
• Authors: Amar Belghit, Mourad Lazri, Ali Hamroun, Karim Labadi, Sadjia Hamdad
• DOI: 10.1051/e3sconf/202669903006/pdf

Research Groups

Not explicitly stated in the provided text.

Short Summary

This study implements and evaluates an AdaBoost-enhanced multiclass One-versus-All Support Vector Machine (AdaOvA-SVM) model for classifying and delineating precipitating clouds in northern Algeria using satellite and radar data, demonstrating its superior performance compared to existing techniques.

Objective

• To implement and evaluate the AdaBoost algorithm to optimize and enhance the performance of multiclass One-versus-All Support Vector Machine (OvA-SVM) for the classification and delineation of precipitating clouds in northern Algeria using MSG-SEVIRI satellite data.

Study Configuration

• Spatial Scale: Northern Algeria, with specific radar data from Sétif.
• Temporal Scale: Not explicitly stated, but involves training and testing phases using satellite images and radar data.

Methodology and Data

• Models used: AdaBoost algorithm, multiclass One-versus-All Support Vector Machine (OvA-SVM), Convective/Stratiform Rain Area Delineation Technique (CS-RADT), Random Forest Technique (RFT). The primary model developed is AdaBoost with OvA-SVM (AdaOvA-SVM).
• Data sources: MSG-SEVIRI (Meteosat Second Generation-Spinning Enhanced Visible and Infrared Imaging) satellite images, Sétif meteorological Radar data (used for training, testing, and validation).

Main Results

• The AdaBoost with OvA-SVM (AdaOvA-SVM) model achieved high performance in precipitating cloud classification.
• Evaluation parameters for AdaOvA-SVM were: Probability of Detection (POD) = 95.2%, Probability of False Detection (POFD) = 12.4%, False Alarm Ratio (FAR) = 14.7%, BIAS = 0.9, Critical Success Index (CSI) = 88.1%, and Percentage Correct (PC) = 96.5%.
• The AdaOvA-SVM technique significantly outperformed the Convective/Stratiform Rain Area Delineation Technique (CS-RADT) and the Random Forest Technique (RFT) in cloud classification performances.
• AdaBoost was shown to effectively improve and optimize the classification accuracy of the multiclass OvA-SVM used as its weak classifier.

Contributions

• Demonstrated the effective application of the AdaBoost algorithm to significantly enhance the performance of multiclass OvA-SVM for precipitating cloud delineation from meteorological satellite data.
• Developed and validated a novel AdaBoost-OvA-SVM (AdaOvA-SVM) model that surpasses established techniques (CS-RADT, RFT) in classification accuracy for the study region of northern Algeria.
• Provided a robust machine learning approach for improving precipitation retrieval and cloud classification, particularly valuable for regions benefiting from enhanced satellite data analysis.

Funding

Not explicitly stated in the provided text.

Citation

@article{Belghit2026Applying,
  author = {Belghit, Amar and Lazri, Mourad and Hamroun, Ali and Labadi, Karim and Hamdad, Sadjia},
  title = {Applying AdaBoost algorithm on multiclass OvA-SVM for the delineation of rainy clouds using multispectral MSG-SEVIRI data},
  journal = {Springer Link (Chiba Institute of Technology)},
  year = {2026},
  doi = {10.1051/e3sconf/202669903006/pdf},
  url = {https://doi.org/10.1051/e3sconf/202669903006/pdf}
}