Sensitivity Analysis Method for Geographical Environmental Changes Based on Support Vector Machines

This study develops a sensitivity analysis method for geographical environmental changes in a typical region of the Yangtze River’s middle reaches. A support vector machine (SVM) model was used to assess the influence of factors such as slope, precipitation, and land use on environmental sensitivity. Key factors were screened and weighted using principal component analysis and recursive feature elimination. Model parameters were optimized with grid search and Bayesian techniques. Results highlight land use intensity and slope as primary drivers of regional sensitivity, with urbanization notably impacting spatial sensitivity distribution. Cross-validation demonstrated strong model accuracy and applicability for environmental sensitivity prediction.