Uncovering controlling factors on rock glacier velocities in the Pamir-Karakoram-Kunlun region using explainable machine learning

This dataset includes the data and codes used in the study "Uncovering controlling factors on rock glacier velocities in the Pamir-Karakoram-Kunlun region using explainable machine learning". Below are the content of each directory in this dataset: 1. data_preparation PKK_RoG.xlsx: attribute table of the analyzed rock glaciers in the Pamir-Karakoram-Kunlun region data_preparation.ipynb: code for forming modeling samples from PKK_RoG.xlsx data: modeling samples, output from data_preparation.ipynb 2. topoclimate_characteristics plot_topoclimate.ipynb: codes for plotting the topographic and climatic distributions of glacier- and talus-connected rock glaciers 3. machine_learning_modeling SVM.ipynb: codes for performing machine learning modeling using Support Vector Machine XGBoost.ipynb: codes for performing machine learning modeling using Extreme Gradient Boost RF.ipynb: codes performing machine learning modeling using Random Forest BPNN.ipynb: codes for performing machine learning modeling using Back Propagation Neural Network 4. SHAP_analysis SVM_SHAP.ipynb: codes for performing SHAP analysis based on the output from Support Vector Machine model XGBoost_SHAP.ipynb: codes for performing SHAP analysis based on the output from Extreme Gradient Boost model RF_SHAP.ipynb: codes for performing SHAP analysis based on the output from Random Forest model BPNN_SHAP.ipynb: codes for performing SHAP analysis based on the output from Back Propagation Neural Network model本数据集涵盖了题为《利用可解释机器学习（explainable machine learning）揭示帕米尔-喀喇昆仑-昆仑地区岩质冰川（rock glacier）速度控制因子》的研究中所使用的数据与代码。以下为本数据集各子目录的内容： 1. data_preparation PKK_RoG.xlsx：帕米尔-喀喇昆仑-昆仑地区已分析岩质冰川的属性表 data_preparation.ipynb：用于从PKK_RoG.xlsx生成建模样本的代码 data：建模样本文件，由data_preparation.ipynb输出得到 2. topoclimate_characteristics plot_topoclimate.ipynb：用于绘制冰川连接型与岩屑堆型岩质冰川的地形及气候分布特征的代码 3. machine_learning_modeling SVM.ipynb：基于支持向量机（Support Vector Machine，SVM）开展机器学习建模的代码 XGBoost.ipynb：基于极限梯度提升（Extreme Gradient Boost，XGBoost）开展机器学习建模的代码 RF.ipynb：基于随机森林（Random Forest，RF）开展机器学习建模的代码 BPNN.ipynb：基于反向传播神经网络（Back Propagation Neural Network，BPNN）开展机器学习建模的代码 4. SHAP_analysis SVM_SHAP.ipynb：基于支持向量机模型输出结果开展SHAP分析（SHapley Additive exPlanations，SHAP）的代码 XGBoost_SHAP.ipynb：基于极限梯度提升模型输出结果开展SHAP分析的代码 RF_SHAP.ipynb：基于随机森林模型输出结果开展SHAP分析的代码 BPNN_SHAP.ipynb：基于反向传播神经网络模型输出结果开展SHAP分析的代码