Applying AdaBoost algorithm on multiclass OvA-SVM for the delineation of rainy clouds using multispectral MSG-SEVIRI data

The use of artificial intelligence and machine learning methods has become a very useful and efficient choice in precipitation retrieval from meteorological satellite data. In this work, we implement the AdaBoost algorithm to optimize and enhance the performance of the classification and delineation of precipitating clouds in northern Algeria carried out by multiclass One-versus-All Support Vector Machine (OvA-SVM). The model developed which combines the AdaBoost algorithm with a multiclass OvA-SVM is applied to images from the MSG-SEVIRI (Meteosat Second Generation-Spinning Enhanced Visible and Infrared Imaging) satellite, with Sétif meteorological Radar data for training and testing validation phases, in which we also did the tuning for setting the adequate number of iterations to stop the AdaBoost ensemble algorithm. In order to evaluate the elaborated model, two classification techniques used previously for rainy clouds delineation in our study region, namely the Convective/Stratiform Rain Area Delineation Technique (CS-RADT) and the Random Forest technique (RFT) are applied for comparison with our built model. The classification results obtained show that AdaBoost with OvA-SVM (AdaOvA-SVM) presents very interesting performances where the evaluation parameters POD, POFD, FAR, BIAS, CSI and PC indicate the values 95.2%, 12.4%, 14.7%, 0.9, 88.1% and 96.5% respectively. Indeed, the AdaOvA-SVM technique has outperformed the CS-RADT and RFT techniques showing better cloud classification performances. At the end of this study, it is shown that the AdaBoost can improve and optimize the classification accuracy of the multiclass OvA-SVM used as its weak classifier.