Estimating construction duration of diaphragm wall using firefly-tuned least squares support vector machine

Neural Computing and Applications - Diaphragm wall is a widely used method for excavating the foundation of buildings. Foundation construction plays a prominent role since it is a predecessor of...     

Enhancing the accuracy of rainfall-induced landslide prediction along mountain roads with a GIS-based random forest classifier

Along mountain roads, rainfall-triggered landslides are typical disasters that cause significant human casualties. Thus, to establish effective mitigation measures, it would be very useful were government agencies and practicing land-use planners to have the capability to make an accurate landslide evaluation. Here, we propose a machine learning methodology for the spatial prediction of rainfall-induced landslides along mountain roads which is based on a random forest classifier (RFC) and a GIS-based dataset. The RFC is used as a supervised learning technique to generalize the classification boundary that separates the input information of ten landslide conditioning factors (slope, aspect, relief amplitude, toposhape, topographic wetness index, distance to roads, distance to rivers, lithology, distance to faults, and rainfall) into two distinctive class labels: ‘landslide’ and ‘non-landslide’. Experimental results with a cross validation process and sensitivity analysis on the RFC model parameters reveal that the proposed model achieves a superior prediction accuracy with an area under the curve  of 0.92. The RFC significantly outperforms other benchmarking methods, including discriminant analysis, logistic regression, artificial neural networks, relevance vector machines, and support vector machines. Based on our experimental outcome and comparative analysis, we strongly recommend the RFC as a very capable tool for spatial modeling of rainfall-induced landslides.

     

Estimation of algal colonization growth on mortar surface using a hybridization of machine learning and metaheuristic optimization

Estimation of the algal colonization growth on façade structure can provide useful information for the task of building maintenance. This research proposes a machine learning method based on the least squares support vector regression (LS-SVR) for modelling the growth time of the green alga Klebsormidium flaccidum on mortar surfaces. Furthermore, to identify an appropriate set of the LS-SVR hyper-parameters, the flower pollination algorithm (FPA) is employed as an optimization technique. The characteristics of the mortar samples, including surface roughness, porosity, surface pH, carbonated condition and type of cement, are employed as input factors for the analysing process. This study relies on a dataset that records 539 laboratory experiments to establish a hybrid model of the LS-SVR and the FPA. The cross-validation process reveals that the proposed method can successfully capture the functional relationship between the algal colonization growth and its influencing factors with a satisfactory outcome (the coefficient of determination R ² = 0.94 and the root mean square error RMSE = 4.55). These facts demonstrate that the hybrid model is a promising tool for assisting the decision-making process in building maintenance planning.     

A swarm intelligence-based machine learning approach for predicting soil shear strength for road construction: a case study at Trung Luong National Expressway Project (Vietnam)

Determining the shear strength of soil is an important task in the design phase of construction project. This study puts forward an artificial intelligence (AI) solution to estimate this parameter of soil. The proposed approach is a hybrid AI model that integrates the least squares support vector machine (LSSVM) and the cuckoo search optimization (CSO). A dataset of 332 soil samples collected from the Trung Luong National Expressway Project in Viet Nam have been used for constructing and validating the AI model. The sample depth, sand percentage, loam percentage, clay percentage, moisture content, wet density of soil, specific gravity, liquid limit, plastic limit, plastic index, and liquid index are used as input variables to predict the output variable of shear strength. In the hybrid AI framework, LSSVM is employed to generalize the functional mapping that estimates the shear strength from the information provided by the aforementioned input variables. Since the model establishment of LSSVM requires a proper setting of the regularization and the kernel function parameters, the CSO algorithm is utilized to automatically determine these parameters. Experimental results show that the prediction accuracy of the hybrid method of LSSVM and CSO (RMSE = 0.082, MAPE = 14.841, and R² = 0.885) is better than those of the benchmark approaches including the standard LSSVM, the artificial neural network, and the regression tree. Therefore, the proposed method is a promising alternative for assisting construction engineers in the task of soil shear strength estimation.

     

Predicting earthquake-induced soil liquefaction based on a hybridization of kernel Fisher discriminant analysis and a least squares support vector machine: a multi-dataset study

Assessment of the earthquake-induced liquefaction potential is a critical concern in design processes of construction projects. This study proposes a novel soft computing model with a hierarchical structure for evaluating earthquake-induced soil liquefaction. The new approach, named KFDA-LSSVM, combines kernel Fisher discriminant analysis (KFDA) with a least squares support vector machine (LSSVM). Based on the original data set, KFDA is used as a first-level analysis to construct an additional feature that best represents the data structure with consideration of different class labels. In the next level of analysis, based on such additional features and the original features, LSSVM generalizes a classification boundary that separates the learning space into two decision domains: “liquefaction” and “non-liquefaction.” Three data sets of liquefaction records have been used to train and verify the proposed method. The model performance is reliably assessed via a repeated sub-sampling process. Experimental results supported by the Wilcoxon signed-rank test demonstrate significant improvements of the hybrid framework over other benchmark approaches.     

A success history-based adaptive differential evolution optimized support vector regression for estimating plastic viscosity of fresh concrete

Engineering with Computers - Plastic viscosity is an important parameter of fresh concrete mixes. This research investigates a machine learning-based method for constructing a functional mapping...     

A hybrid computational intelligence approach for predicting soil shear strength for urban housing construction: a case study at Vinhomes Imperia project,Hai Phong city (Vietnam)

This research proposes an alternative for estimating shear strength of soil based on a hybridization of Support Vector Regression (SVR) and Particle Swarm Optimization (PSO). SVR is used as a function approximation method for making prediction of the soil shear strength based on a set of twelve variables including sample depth, sand content, loam content clay content, moisture content, wet density, dry density, void ratio, liquid limit, plastic limit, plastic index, and liquid index. The hybrid framework, named as PSO–SVR, relies on PSO, as a metaheuristic, to optimize the training phase of the employed function approximator. A data set consisting of 443 soil samples associated with the experimental results of shear strength has been collected from a housing project in Vietnam. This data set is then used to train and verify the performance of the PSO–SVR model specifically constructed for shear strength estimation. The hybrid model has achieved a good modeling outcome with Root Mean Square Error (RMSE) = 0.038, Mean Absolute Percentage Error (MAPE) = 9.701%, and Coefficient of Determination (R2) = 0.888. Hence, the PSO–SVR model can be a potential alternative to be participated in the design phase of high-rise housing projects.     

Prediction of long-term deflections of reinforced-concrete members using a novel swarm optimized extreme gradient boosting machine

During the life cycle of buildings and infrastructure systems, the deflection of reinforced-concrete members generally increases due to both internal and external factors. Accurate forecasting of long-term deflection of these members can significantly enhance the effectiveness of structural maintenance processes. This research develops a hybrid data-driven method which employs the extreme gradient boosting machine and the particle swarm optimization metaheuristic for predicting long-term deflections of reinforced-concrete members. The former, a machine learning technique, generalizes a non-linear mapping function that helps to infer long-term deflection results from the input data. The later, a swarm-based metaheuristic, aims at optimizing the machine learning model by fine-tuning its hyper-parameters. The proposed hybridization of machine learning and swarm intelligence is constructed and verified by a dataset consisting of 217 experiments. The experiment results, supported by statistical tests, point out that the hybrid framework is able to attain good predictive performances with average root-mean-square error of 11.38 (a reduction of 17.4%), and average coefficient of determination of 0.88 (an increase of 6.0%) compared to the non-hybrid model. These results also outperform those obtained by other popular techniques, including Backpropagation Neural Networks and Regression Tree in several popular benchmarks, such as root-mean-square error, mean absolute percentage error, and the coefficient of determination R². This is backed up by statistical tests with the level of significance \(\alpha = 0.05\). Therefore, the newly developed model can be a promising tool to assist civil engineers in forecasting deflections of reinforced-concrete members.

     

A novel time-depended evolutionary fuzzy SVM inference model for estimating construction project at completion

Construction projects frequently face cost overruns during the construction phase. Thus, a proactive approach is essential for monitoring project costs and detection of potential problems. In construction management, Estimate at Completion (EAC) is an indicator for assisting project managers in identifying potential problems and developing appropriate responses. This study utilizes weighted Support Vector Machine (wSVM), fuzzy logic, and fast messy Genetic Algorithm (fmGA) to handle distinct characteristics in EAC prediction. The wSVM is employed as a supervised learning technique that can address the features of time series data. The fuzzy logic is aimed to enhance the model capability of approximate reasoning and to deal with uncertainty in EAC prediction. Moreover, fmGA is utilized to optimize model's tuning parameters. Simulation results show that the new developed model has achieved a significant improvement in EAC forecasting.