Accurate assessment of undrained shear strength (USS) for soft sensitive clays is a great concern in geotechnical engineering practice. This study applies novel data-driven extreme gradient boosting (XGBoost) and random forest (RF) ensemble learning methods capturing the relationships between USS various basic soil parameters. Based on data sets from TC304 database, general approach developed to predict using two machine above, where five feature variables including preconsolidation stress...

Driven pile is widely used as an effective and convenient structural component to transfer superstructure loads deep stiffer soils. Nevertheless, during the design process of piles, due intrinsic complexity well various variables, internal stress state related drivability remains unclear, which makes analysis imprecise. Thus, development accurate predictive model becomes emergent. This paper presents a practical approach assess in relation prediction Maximum compressive stresses Blow per...

Historically, landslides have been the primary type of geological disaster worldwide. Generally, stability reservoir banks is primarily affected by rainfall and water level fluctuations. Moreover, changes with long-term dynamics external disaster-causing factors. Thus, assessing time-varying reliability remains a challenge. In this paper, machine learning (ML) based approach proposed to analyze bank in spatially variable soils through time series prediction. This study systematically...

Embankments are widespread throughout the world and their safety under seismic conditions is a primary concern in geotechnical engineering community since failure events may lead to disastrous consequences. This study proposes an efficient slope stability analysis approach by introducing advanced gradient boosting algorithms, namely Categorical Boosting (CatBoost), Light Gradient Machine (LightGBM), Extreme (XGBoost). A database consisting of 600 datasets prepared for model calibration...

Rockhead profile is an important part of geological profiles and can have significant impacts on some geotechnical engineering practice, thus, it necessary to establish a useful method reverse the rockhead using site investigation results. As general reflect spatial distribution geo-material properties based field measurements, conditional random (CRF) was improved in this paper simulate profiles. Besides, measurements are generally limited due limitations budget time so that estimation mean...

Determining soil–water characteristic curve (SWCC) at a site is an essential step for implementing unsaturated soil mechanics in geotechnical engineering practice, which can be measured directly through various in-situ and/or laboratory tests. Such direct measurements are, however, costly and time-consuming due to high standards equipment procedural control limits testing apparatus. As result, only limited number of data points (e.g., volumetric water content vs. matric suction) on SWCC some...

Promoting customers' pro-environmental behaviors (PEBs) is crucial for greening the hotel industry. However, customers often hesitate to adopt resource-saving PEBs in public places like hotels due a "dual positive externality/spillover" problem – extra economic spillovers and increased environmental spillovers, compared conducting such private households. This study explores habitual at home two distinct cognitions associated with key spillover beneficiaries (nature hotels) address this...

Reliability analysis plays an important role in the risk management of geotechnical engineering. For random field-based method, it is expected that uncertainty characterization geo-material parameters and realization field can be integrated effectively. Moreover, as increase measured data size generally difficult investigation engineering due to limitation budget time etc., statistical resulting from sparse should paid great attention. Therefore, taking determination hyper-parameters for...

Generation of particles with irregular shape and the overlap detection are crucial for numerical simulation granular materials. This paper presents a systematic approach to develop two-dimensional random particle model Firstly, angular bend (RAB) algorithm is proposed coded in Python simulate geometric individual shape. Three representative parameters used quantitatively control feature generated polygons terms three major aspects, respectively. Then, geometrical models implemented into flow...