Investigating soil grouping using conventional and modern clustering models in some parts of Qazvin plain
Author(s):
Article Type:
Research/Original Article (دارای رتبه معتبر)
Abstract:
Soil is a crucial component in achieving sustainable development goals due to its significant role in addressing environmental challenges. It is essential to differentiate soils that have similar management requirements. This necessity has prompted soil scientists to employ numerical classification models to categorize soils based on their similarities. In this study, we utilized two types of clustering models, traditional and modern, to classify soils from certain areas of the Qazvin Plain. Using one-way and two-way clustering models, we grouped 297 soils from the region based on a comprehensive set of their morphological, physicochemical, and environmental attributes. The classifications derived from these two models were assessed using internal and external evaluation indicators, with the distribution map of soil subgroups serving as a ground truth reference map. The results indicated that the hierarchical clustering model, with a lower Davis-Bouldin index (DB: 1.38) and a higher adjusted Rand index (ARI: 0.49), outperformed the biclustering model. However, the classifications from the bidirectional clustering model corresponded reasonably well with the topographical and soil changes in the region, as evidenced by the higher Shannon’s difference index in the bidirectional clustering model (1.82) compared to the hierarchical clustering model (1.62). Overall, the study’s findings underscore the utility of the co-clustering model as a contemporary data mining technique for soil classification and identification of soil management similarity patterns.
Keywords:
Language:
Persian
Published:
Iranian Journal of Soil and Water Research, Volume:55 Issue: 8, 2024
Pages:
1273 to 1295
https://www.magiran.com/p2787016
سامانه نویسندگان
مقالات دیگری از این نویسنده (گان)
-
Uncertainty and Spatial mapping of soil salinity and sodicity using machine learning methods in three different management depths in Abyek region
Azam Jafari, *, Zahra Rasaei
Iranian Journal of Soil and Water Research, Jun 2025 -
Assessment of environmental indices for soil lead contamination in a part of Shahr-e-Ray, Tehran Province
Shahla Rahmani Siyalarz, Ali Keshavarzi *, , Mohsen Farahbakhsh
Iranian Journal of Soil and Water Research,