Red Sandstone Dynamic Compressive Strength Under Acidic Wetting-Drying Cycles

A dataset of 597 experimental test results on red sandstone samples, investigating dynamic compressive strength under coupled acidic wetting-drying cycles. Bin Du authored this dataset, which was last updated on May 14, 2026. It was used to develop and validate five ensemble learning models for predicting mechanical behavior in chemically degraded environments.

Use Cases

• Training ensemble learning models for rock strength prediction based on strain rate and environmental cycle data
• Interpreting feature importance in geotechnical models using SHAP analysis based on the five input variables
• Developing graphical user interfaces for intuitive prediction of dynamic compressive strength based on the XGBoost model
• Comparing the performance of Random Forest, AdaBoost, XGBoost, LightGBM, and GBDT models on experimental geotechnical data

Strengths

• Contains 597 experimental test results, providing a substantial basis for model training
• Evaluates five distinct ensemble learning models (RF, AdaBoost, XGBoost, LightGBM, GBDT) with four performance metrics (R², RMSE, MAE, MAPE)
• Includes SHAP analysis for model interpretability, identifying strain rate as the most influential factor

Limitations

• Column-level documentation is absent; field semantics must be inferred after download
• Row count is unknown, which may limit suitability assessment
• Data may reflect bias inherent to the specific experimental conditions and red sandstone samples used

Provenance

Source

figshare

Collection Method

Laboratory experiments on red sandstone samples under coupled acidic wetting-drying cycles

Freshness

Last updated 2026-05-14 17:26:51; freshness should be verified