A One-Class-Classifier-Based Negative Data Generation Method for Rapid Earthquake-Induced Landslide Susceptibility Mapping

Abstract

Machine learning with extensively labeled training samples (e.g., positive and negative data) has received much attention in terms of addressing earthquake-induced landslide susceptibility mapping (LSM). However, the extensive amount of labeled training data required by machine learning, particularly the precise negative data (i.e., non-landslide area), cannot be easily and efficiently collected. To address this issue, this study presents a one-class-classifier-based negative data generation method for rapid earthquake-induced LSM. First, an incomplete landslide inventory (i.e., positive data) was produced with the aid of change detection using before-and-after satellite images and the Geographic Information System (GIS). Second, a one-class classifier was utilized to compute the probability of landslide occurrence based on the incomplete landslide inventory followed by the negative data generation from the low landslide susceptibility areas. Third, the positive data as well as the generated negative data (i.e., non-landslide) were compounded to train a traditional binary classifier to produce the final LSM. Experimental results suggest that the proposed method is capable of achieving a result that is comparable to methods using the complete landslide inventory, and it displays good correspondence with recent landslide events, making it a suitable method for rapid earthquake-induced LSM. The findings in this study would be useful in regional disaster planning and risk reduction.

Ground-Truth based Performance Assessment of Landslide Susceptibility Mapping: A Case Study after Elazig Earthquake (24 Jan 2020, Mw 6.8), Turkey

“…When preparing susceptibility map of landslides triggered by earthquakes, Wang et al (2021) employed elevation, aspect, slope angle, terrain relief, distance to faults, distance to rivers and terrain wetness index as conditioning factors while considered elevation, slope, aspect, land cover, soil type, precipitation, distance to faults, distance to roads and distance to streams for investigating eartquakes on landslides susceptibility in a seismic prone area in Central Asia. Several studies (Chen et al 2021;Guo et al 2021;Liu et al 2021) considered similar conditioning and triggering parameters when producing susceptibility maps of landslides triggered by earthquakes. Consequently, different conditioning factors such as altitude, slope, aspect, plan and pro le curvature, lithology, topographic wetness index (TWI), stream power index (SPI), normalized difference vegetation index (NDVI), drainage density, distance to roads were used to produce LSM in the literature (Chen et al 2017b;de Oliveira et al 2019;Wang et al 2020;Adnan et al 2020;Bui et al 2020).…”

“…In the recent literature, it is possible to nd studies on regional susceptibility assessments of landslides triggered by earthquakes (e.g. Xie et al 2018;Chen et al 2020 a, b, c;Chen et al 2021). However, if there is no landslide inventory prepared immediately after earthquake, it is still di cult to distinguish the landslides in a region that are triggered by earthquakes.…”

“…Algorithms such as arti cial neural networks (ANNs), support vector machine (SVM), decision tree (DT), random forest (RF), gradient boosting machine (GBM), fuzzy logic (FL) and logistic regression (LR) are among the ML algorithms used to produce LSMs (e.g. see Pourghasemi et al 2012;Wang et al 2016;Sevgen et al 2019;Wang et al 2020;Yanar et al 2020;Qi et al 2021;Wu et al 2021). Wang et al (2020) compared susceptibility results for landslide prediction using ve methods such as LR (0.77), SVM (0.80), RF (0.82), GBM (0.81) and MLP (0.80) in Shexian County.…”

Landslide Susceptibility Mapping Based on Interpretable Machine Learning from the Perspective of Geomorphological Differentiation

“…The selection of reasonable negative samples has a significant impact on the prediction results. In previous studies, negative samples have often been selected using environmental similarity-based sampling (ESBS), buffer-controlled sampling (BCS) [39], target space exteriorization sampling (TSES) [40], etc. A-Xing Zhu [41] proposed a negative sample sampling similarity theory that quantifies negative samples based on the environmental similarity between alternative negative and positive samples, and compared it with two existing negative sample generation methods, BCS and TSES.…”

Generating accurate negative samples for landslide susceptibility mapping: A combined self-organizing-map and one-class SVM method

“…, M}, x i ∈ R n . The OC-SVM maps the samples from low-dimensional to high-dimensional space through the Φ, establishing an optimal hyperplane between the zero point and the high-dimensional space (Chen et al, 2021). This hyperplane can be described by the n-dimensional vector W [w 1 , w 2 , .…”