Machine learning modelling for predicting soil liquefaction susceptibility

Abstract

Abstract. This study describes two machine learning techniques applied to predict liquefaction susceptibility of soil based on the standard penetration test (SPT) data from the 1999 Chi-Chi, Taiwan earthquake. The first machine learning technique which uses Artificial Neural Network (ANN) based on multi-layer perceptions (MLP) that are trained with Levenberg-Marquardt backpropagation algorithm. The second machine learning technique uses the Support Vector machine (SVM) that is firmly based on the theory of statistical learning theory, uses classification technique. ANN and SVM have been developed to predict liquefaction susceptibility using corrected SPT [(N 1 ) 60 ] and cyclic stress ratio (CSR). Further, an attempt has been made to simplify the models, requiring only the two parameters [(N 1 ) 60 and peck ground acceleration (a max /g)], for the prediction of liquefaction susceptibility. The developed ANN and SVM models have also been applied to different case histories available globally. The paper also highlights the capability of the SVM over the ANN models.

Toward reliable storm-hazard forecasts: XBeach calibration and its potential application in an operational early-warning system

“…Tides in the area are semi-diurnal, with average range of 2.8 m for spring-tides and 1.3 m during neap tides, although a maximum range of 3.5 m can be reached (e.g., Ferreira et al 2006 is moderate to high, with an average annual significant offshore wave height H s 00.92 m and average peak wave period T p 08.2 s (Ferreira et al 2009). Waves are mostly west-southwest (occurrence 71%), while shorter-period SE waves generated by regional winds are also frequent (23%) (Almeida et al 2011c). Storm events in the region are considered as those in which the significant offshore wave height exceeds 3 m; such events typically correspond to less than 2% of the offshore wave climate (Almeida et al 2011c).…”

“…Waves are mostly west-southwest (occurrence 71%), while shorter-period SE waves generated by regional winds are also frequent (23%) (Almeida et al 2011c). Storm events in the region are considered as those in which the significant offshore wave height exceeds 3 m; such events typically correspond to less than 2% of the offshore wave climate (Almeida et al 2011c). Faro Beach is characterized by a steep beach-face, with an average slope of around 10% and varying from 6% to 15% (Vousdoukas et al , 2011b.…”

Bioclimatic analysis in a region of southern Italy (Calabria)

“…In particular, several researches have been carried out to investigate trends in annual and seasonal precipitation, at a large scale (Kutiel et al 1996;Piervitali et al 1998;Xoplaki et al 2006), and for entire nations or regions (Esteban-Parra et al 1998;De Luis et al 2000;Feidas et al 2007;Río et al 2011). Long precipitation records have been investigated in northern and central Italy (Montanari et al 1996;Demichele et al 1998;Brunetti et al 2006b), in southern Italy (Palmieri et al 1991;Brunetti et al 2004;Brunetti et al 2006a;Samui et al 2011b) and, particularly, in the Calabria Region (Coscarelli et al 2004;Buttafuoco et al 2011aButtafuoco et al , 2011bCaloiero et al 2011a;Brunetti et al 2012;Ferrari et al 2013;Caloiero et al 2014;Sirangelo et al 2015). With respect to the temperature, the majority of these studies have been conducted at large spatial scale (Easterling et al 2000;Klein Tank & Können 2003;Vose et al 2004;Vincent et al 2005) or at national spatial scale (Domonkos & Tar 2003;Brunetti et al 2006a), while few studies have been made at local scale (Brunetti et al 2004;Piccarreta et al 2004;Buttafuoco et al 2010;Caloiero et al 2014).…”

Prediction of pore-water pressure response to rainfall using support vector regression

“…Support vector machines have also been applied to geotechnical engineering problems-the study by Goh and Goh (2007) on seismic liquefaction data is one of the early applications of SVM in geotechnical engineering. Samui and Sitharam (2011) used SVM to classify expansive soils; Yu et al (2012) used SVM along with data assimilation techniques to predict soil moisture both at the surface and the root zone; Gill et al (2006) performed 4 and 7-day forecasts of soil moisture while exploiting the relationship between the soil moisture and metrological factors; Samui and Karthikeyan (2011) predicted the susceptibility to liquefaction of soils using features of cone resistance and cyclic stress ratio, in which the model was locally trained and tested and even further extended to a global data set with reasonable success. Lee and Chern (2013) used SVM to classify liquefied and non-liquefied soils.…”