Summary
Effective modelling and integrated spectral analysis approaches can advance modelling precision. To develop an integrated spectral forecast modelling of soil organic carbon (SOC), this research investigated a mining coal in Dengcao Coal Mine Area, Zhengzhou. The study utilizes the Lasso and Ranger algorithms were utilized in spectral band analysis. Four primary models employed during this process include Artificial Neural Network (ANN), Support Vector Machine, Random Forest (RF), and Partial Least Squares Regression (PLSR). The ideal model was chosen. The results showed that, in contrast to when band collection was based on Lasso algorithm modelling, model precision was higher when it was based on the Ranger algorithm. ANN model had an ideal goodness acceptance, and the modelling developed
Regional applicability
This study is location-specific to Chinese coal mining areas and the methodology's transferability to United Kingdom mining contexts would depend on soil type, spectral characteristics, and climatic conditions. The UK has fewer active deep coal mines but extensive legacy mining areas; the spectral modelling approach could be adapted for UK mine site remediation and monitoring programmes, though validation would be required.
Key measures
Soil organic carbon (SOC) content predicted via spectral analysis; model precision comparison across Lasso vs. Ranger band selection algorithms; performance metrics for ANN, SVM, RF, and PLSR models
Outcomes reported
The study developed and compared four machine learning models (ANN, SVM, RF, PLSR) integrated with spectral analysis to predict soil organic carbon content in coal-mined areas. The Ranger algorithm for spectral band selection outperformed Lasso, with Random Forest showing the most stable predictions.
Topic tags
Dig deeper with Pulse AI.
Pulse AI has read the whole catalogue. Ask about this record, its theme, or how the findings apply to UK farming and policy — every answer cites the underlying studies.