Fig 1.
Overview of study area and schematic diagram of sampling locations.
(a)General distribution map of loess in Ili region (Schematic redrawn based on geological information from Ye et al. [20]); (b)Typical geological cross-section of loess strata; (c)Field photograph of Upper Pleistocene aeolian loess on slope surface; (d)Field photograph of Holocene colluvial loess at slope toe.
Table 1.
Probabilistic statistical analysis results of loess geotechnical parameters.
Table 2.
Statistical results of correlation analysis between loess compression coefficient and soil properties.
Table 3.
Statistical table of selected soil property correlation coefficients.
Fig 2.
Scatter plots of loess compression coefficient vs. compression modulus correlation analysis.
(a) Huocheng County; (b) Nilka County; (c) Xinyuan County.
Fig 3.
Scatter plots of loess compression coefficient vs. void ratio correlation analysis.
(a) Huocheng County; (b) Nilka County; (c) Xinyuan County.
Fig 4.
Scatter plots of loess compression coefficient vs. dry density correlation analysis.
(a) Huocheng County; (b) Nilka County; (c) Xinyuan County.
Table 4.
Summary of loess compressibility and soil property regression models.
Table 5.
Regression coefficients and significance analysis for loess compressibility and soil properties.
Table 6.
Comparison of loess compressibility multiple regression model prediction results with measured data.
Fig 5.
Error Plots of Loess Compressibility Multiple Regression Model Predictions vs. Measured Values.
(a) Huocheng County; (b) Nilka County; (c) Xinyuan County.
Table 7.
Dataset partitioning for random forest prediction model.
Table 8.
Comparison of loess compressibility random forest model prediction results with measured data.
Fig 6.
Error Plots of Loess Compressibility Random Forest Model Predictions vs. Measured Values.
(a) Huocheng County; (b) Nilka County; (c) Xinyuan County.
Table 9.
Dataset partitioning for multilayer perceptron neural network prediction model.
Table 10.
Comparison of loess compressibility multilayer perceptron neural network model prediction results with measured data.
Fig 7.
Error Plots of Loess Compressibility Multilayer Perceptron Neural Network Model Predictions vs. Measured Values.
(a) Huocheng County; (b) Nilka County; (c) Xinyuan County.
Fig 8.
Scanning Electron Microscopy (SEM) images of Ili loess microstructure.
(a) ×1500; (b) ×6000; (c) ×30000.
Fig 9.
Comparison of Measured Compression Coefficient Values with Predicted Values from the Established Model.
(a) Huocheng County; (b) Nilka County; (c) Xinyuan County.
Table 11.
Comparative statistical table of established prediction models for each region.