Comparative landslide spatial research based on various sample sizes and ratios in Penang Island, Malaysia

  • PDF / 4,314,842 Bytes
  • 22 Pages / 595.276 x 790.866 pts Page_size
  • 65 Downloads / 159 Views

DOWNLOAD

REPORT


ORIGINAL PAPER

Comparative landslide spatial research based on various sample sizes and ratios in Penang Island, Malaysia Han Gao 1 & Pei Shan Fam 1 & Lea Tien Tay 2 & Heng Chin Low 3 Received: 24 February 2020 / Accepted: 7 September 2020 # Springer-Verlag GmbH Germany, part of Springer Nature 2020

Abstract This paper aims to compare and develop the influence on different sample sizes and sample ratios when using machine learning (ML) models, i.e., support vector machine (SVM) and artificial neural network (ANN), to produce landslide susceptibility maps (LSMs) in Penang Island, Malaysia. At the same time, traditional statistical (TS) models are also considered to produce LSMs in this comparative research. The receiver operating characteristic (ROC) curve and recall metric are applied to evaluate the model’s performance. Based on the evaluation criteria, the ML model outperforms the TS models and the ML models trained using the datasets with larger sample size give a better performance. ML models, especially SVM models, have better performance when training with balanced datasets as well as the datasets of more landslide sample data. Kruskal-Wallis test and Mann-Whitney U test are applied to test the significance. The results indicate that sample size and sample ratio are essential factors when considering ML models to produce LSMs. The LSMs produced in this research can provide valid and useful information to the local authorities for landslide mitigation and prediction. Keywords Landslide susceptibility mapping . Artificial neural network . Support vector machine . Sample size . Sample ratio . Receiver operating characteristic

Introduction Landslides cause great losses of lives and properties around the world every year (Scaringi et al. 2018). Petley (2012) indicated that 2620 fatal landslides in total were recorded

* Pei Shan Fam [email protected] Han Gao [email protected] Lea Tien Tay [email protected] Heng Chin Low [email protected] 1

School of Mathematical Sciences, Universiti Sains Malaysia, 11800 USM, Penang, Malaysia

2

School of Electrical and Electronic Engineering, USM, Engineering Campus, Seberang Perai Selatan, 14300 Nibong Tebal, Penang, Malaysia

3

Research and Innovation Unit, Universiti Sains Malaysia, 11800 USM, Penang, Malaysia

worldwide between 2004 and 2010, causing a total of 32,322 recorded fatalities. Klose et al. (2015) estimated the transportation infrastructure losses in their research and found that the landslide loss for highways in the USA amounted to a total of USD23.5 million from 1980 to 2010. According to the data from the US National Aeronautics and Space Administration (NASA) (http//www.nasa.gov), Malaysia experienced 171 landslides from 2007 to 2016, which made the country rank one of the top 10 countries in frequency of landslides. Landslide spatial prediction research, such as landslide susceptibility assessment (LSA), plays an essential role in landslide mitigation and management, which enables the authorities to take relative measures to reduce the damage to lives and proper