Dataset for Landslide Susceptibility Prediction

This dataset considers 7 landslide conditioning factors such as Curvature, Slope, Aspect, Elevation, NDVI, Precipitation, and LULC for 100 random locations in Raigad district of Maharashtra, India. These factors were categorized based on their range of values and influence on landslide occurrence, ranging from very high(5), high(4), moderate(3), low(2) to very low(1). Further based on the values of these factors the landslide susceptibility of that location is categorized into low (1), moderate(2) and high(3). Thus this dataset can be used for multiclass classification of landslides using machine learning algorithm.

本数据集针对印度马哈拉施特拉邦赖加德（Raigad）区的100个随机采样点位，纳入曲率（Curvature）、坡度（Slope）、坡向（Aspect）、高程（Elevation）、归一化植被指数（NDVI, Normalized Difference Vegetation Index）、降水量（Precipitation）以及土地利用/覆被（LULC, Land Use/Land Cover）共7类滑坡影响因子。上述因子依据其数值区间及对滑坡发生的影响程度被划分为5个等级，由高至低依次为极高（5）、高（4）、中等（3）、低（2）与极低（1）。进一步基于上述因子的数值，可将各点位的滑坡敏感性划分为低（1）、中等（2）与高（3）三类。因此，本数据集可用于基于机器学习算法的滑坡多分类任务。