ЖЕРДІҢ ТҰЗДЫҒЫН СЕГМЕНТАЦИЯЛАУ МОДЕЛІН ДАЛАЛЫҚ ЗЕРТТЕУЛЕР МЕН КЛИМАТТЫҚ ПАРАМЕТРЛЕР НЕГІЗІНДЕ ӘЗІРЛЕУ

Abstract

The paper presents a hybrid machine learning model for the spatial segmentation of soils by salinity using multispectral satellite data from Sentinel-2 and climate parame-ters of the ERA5-Land model. The proposed method aims to solve the problem of ac-curate soil cover segmentation under climate change and high spatial heterogeneity of data. The approach includes the sequential application of unsupervised learning algo-rithms (KMeans, hierarchical clustering, DBSCAN), the XGBoost model, and a multi-tasking neural network that performs simultaneous classification and regression. At the first stage, pseudo-labels are formed using KMeans, then a probabilistic assessment of object membership in classes and ensemble voting of clustering algorithms are car-ried out. The final model is trained on an extended feature space and demonstrates improved results compared to traditional approaches. Experiments on a sample of 33,624 observations (23,536 — training sample, 10,088 — test sample) showed an in-crease in the Silhouette Score value from 0.7840 to 0.8156 and a decrease in the Da-vies-Bouldin Score from 0.3567 to 0.3022. The classification accuracy was 99.99%, with only one error in more than 10,000 test objects. The results confirmed the proposed method's high efficiency and applicability for remote monitoring, environmental analysis, and sustainable land management.