- Two new modeling methods were added in The LSM_ToolPack namely, Support Vector Machine (SVM) and eXtreme gradient boosting (XGBoost)
- Data Preparation(Train/Validation Split)
- Feature Selection
- Create LSM with Logistic Regression
- Create LSM with Random Forest
- Performance Evaluation
- Create Raster Stack (Multi-Bands)
- Support Vector Machine
- eXtreme gradient boosting (XGBoost)
- ONLY SUPPORTING ArcGIS PRO 2.x (64Bit) and R BASE 4.2x
- ArcGIS Pro 1.1 or later (don't have it? try trial edition)
- R Statistical Computing Software, 4.2.x or later (What is R?). !!! R Base 3.6 is not supported !!! Recommended Version v4.2.x
- ArcGIS R-Bridge
- Recommended Version v1.0.1.300
- NOT: If arcgisbinding is not installed, run this code in R terminal << install.packages("arcgisbinding", repos="https://r.esri.com", type="win.binary") >>
- Java Runtime Environment for FSelector package
- If you are setting up modules for the first time, you will need an internet connection to install the R-packages on the depository.
- First, make sure you've installed the requirements.
- Download this repository and unzip this file
- Open your ArcGIS application
- This clip shows you how to add the toolbox:
This clip shows you how to divide your data train and validation data:
This clip shows you how to use the module for selecting the best feature subset:
This clip shows you how to use the LR algortihm for produce susceptibility map. This module provides the user statistical results and LR model ROC curve and AUC value.
This clip shows you how to use the tool: This module provides the user RF feature importance results as an excel sheet paper and RF model ROC curve and AUC value as a 300dpi TIFF image.
This clip shows you how to use the tool: This module provides the user accuracy metric results (Overall accuracy, Kappa, AUC, and F1 values) as an excel sheet paper.
This clip shows you how to use the tool: this module transforms your single factor maps into raster stack map.
LSM Tool Pack was prepared as part of the projects “Development of ArcGIS Interfaces with R programming language for Landslide Susceptibility Mapping” (No. 118Y090) funded by The Scientific and Technological Research Council of Turkey (TUBITAK).
Emrehan Kutlug Sahin, Ismail Colkesen, Suheda Semih Acmali, Aykut Akgun, Arif Cagdas Aydinoglu, Developing comprehensive geocomputation tools for landslide susceptibility mapping: LSM tool pack, Computers & Geosciences, 2020, 104592, ISSN 0098-3004, https://doi.org/10.1016/j.cageo.2020.104592. (http://www.sciencedirect.com/science/article/pii/S009830042030577X)