Spanish Dataset for Automatic Classification of Ground Deformation Processes

This .rar file contains five files used for training and testing a machine learning algorithm for the automatic classification of ground deformation processes. Four database in XLSX format: MPs.xlsx, 50m.xlsx, 100m.xlsx, and 200m.xlsx. Each file represents a dataset with different scales and contain...

Descripción completa

Detalles Bibliográficos
Autor: Instituto Geológico y Minero de España (IGME)
Tipo de recurso: conjunto de datos
Fecha de publicación:2025
País:España
Institución:Consejo Superior de Investigaciones Científicas (CSIC)
Repositorio:DIGITAL.CSIC. Repositorio Institucional del CSIC
OAI Identifier:oai:digital.csic.es:10261/418947
Acceso en línea:http://hdl.handle.net/10261/418947
Access Level:acceso abierto
Palabra clave:Machine learning
Remote sensing
Radar
Hazard
Subsidence
Landslide
Landslides
Descripción
Sumario:This .rar file contains five files used for training and testing a machine learning algorithm for the automatic classification of ground deformation processes. Four database in XLSX format: MPs.xlsx, 50m.xlsx, 100m.xlsx, and 200m.xlsx. Each file represents a dataset with different scales and contains: • An ID column for each unique record. • A geometry column indicating the coordinates of each record. • A processes column indicating the ground deformation process for each record. The considered processes are: Natural landslide, Mining landslide, Constructive subsidence, Mining subsidence, and Piezometric subsidence • 32 columns, one for each predictive variable, including information about ground deformation, geology, morphometry, climate, and land use. One results file in XLSX format "Hierarchical_Prediction.xlsx" , contains the following columns: • An ID column that uniquely identifies each record. • A Parent_processes column indicating the main ground deformation process for each record (Landslide and Subsidence). • A processes column specifying the specific ground deformation process for each record. • Columns with prediction score and prediction label from each model (parent model, landslide model, and subsidence model). • A prediction column with the final predicted class. One Word file: Contains the full description of each variable included in the CSV files, as well as the interpretation of the IDs assigned to each record. This dataset allows reproducing experiments on the automatic classification of ground deformation processes and serves as a basis for geospatial analysis.