Soil in combination with cement, more commonly referred to as soilcrete, has gained great popularity within the construction sector. To this end, its mechanical properties have to be determined quickly and accurately. Unfortunately, the conventional methods for determining them include lab tests which are rather expensive and error prone. A better solution however comes in the form of machine learning (ML) algorithms which have tremendous potential. Hence, this study sought to analyze how efficient the three algorithms in predicting the uniaxial compressive strength (UCS) of soilcrete. 400 samples of soilcrete were manufactured and analyzed, using two types of soil including clay and limestone along with metakaolin which served as a mineral additive. A total of 80% of the dataset was made use of for training while the remaining 20% served a testing purpose, in addition to the 37 datasets which were specifically designed for evaluation purposes. A Stepwise procedure was completed and a total of 8 parameters were identified including metakaolin and soil type, super plasticizer content, water to binder ratio, shrinkage, binder density and finally ultrasonic velocity. Most of the algorithms were able to achieve satisfactory results however Gaussian process regression (GPR), support vector regression (SVR) and null-space SVR (NuSVR) were able to stand out due to their potential performance. Focusing on the trained models and lab tests that were done, this research managed to establish the proper superplasticizer constitution (1%), water-to-binder ratio (0.4) and metakaolin content (12%) with the goal of achieving the highest UCS value in the provided soilcrete specimens. Furthermore, a graphical user interface (GUI) was created based on the trained ML models. For the civil engineers and researchers who need to estimate the UCS of the soilcrete specimens, this GUI greatly simplifies the process.

This study presents a nonmodel-based machine learning framework for estimating engineering demand parameters (EDPs) of eccentrically braced frames with soil-structure interaction effects. The objective is to estimate residual and peak story drift ratio, peak floor acceleration, and develop fragility curves using traditional regression equations and advanced machine-learning techniques. Correction coefficients are developed to improve prediction accuracy by accounting for soil-structure interaction. A comprehensive database, including incremental dynamic analysis results of 4- and 8-story frames, is developed, consisting of 109,841 data points. The database includes fixed-base models and models with various soil-structure interaction values, subjected to 44 far-field ground motions. Four scenarios are introduced considering various input variables to compare the impact of soil-structure interaction. Findings reveal the effects of soil-structure interaction features on the performance of machine learning algorithms, increasing by up to 17.61% of the coefficient of determination. Utilizing the predicted story drift ratio, two types of fragility curves indicate more precise predictions, emphasizing the impact of soil-structure interaction effects at lower damage levels. A graphical user interface has been developed to predict fragility curves based on various inputs to promote the practical use of machine learning in engineering. Two new 4-story frames are used as case studies, subjected to unseen ground motions to assess the application of trained machine learning algorithms. Prediction errors in input-output scenarios considering soilstructure interaction range from 3% to 18% for new frames. The proposed approach for predicting EDPs is further acknowledged by evaluating a real instrumented five-story steel frame office building.

Soilcrete is an innovative construction material made by combining naturally occurring earth materials with cement. It can be effectively used in areas where other construction materials are not readily available due to financial or environmental reasons since soilcrete is made from readily available natural clay. It can also help to cut down the greenhouse gas emissions from the construction industry by encouraging the use of resources that are locally available. Thus, it is imperative to reliably predict different properties of soilcrete since the accurate determination of these properties is crucial for the widespread use of soilcrete materials. However, the laboratory determination of these properties is subjected to significant time and resource constraints. As a result, this research was undertaken to provide empirical prediction models for the density, shrinkage, and strain of soilcrete mixes using two machine learning algorithms: Gene Expression Programming (GEP) and Extreme Gradient Boosting (XGB). The analysis revealed that XGB-based predictions correlated more with real-life values than GEP having training R2=0.999\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\text{R}}{2}=0.999$$\end{document} for both density and shrinkage prediction and R2=0.944\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\text{R}}{2}=0.944$$\end{document} for strain prediction. Moreover, several explanatory analyses including individual conditional expectation (ICE) analysis and shapely analysis were done on the XGB model which showed that water-to-binder ratio, metakaolin content, and modulus of elasticity are some of the most important variables for forecasting soilcrete materials properties. Furthermore, an interactive graphical user interface (GUI) has been developed for effective utilization in civil engineering industry to forecast these properties of soilcrete materials.

The resilient modulus (MR) of different pavement materials is one of the most important input parameters for the mechanistic-empirical pavement design approach. The dynamic triaxial test is the most often used method for evaluating the MR, although it is expensive, time-consuming, and requires specialized lab facilities. The purpose of this study is to establish a new model based on Long Short-Term Memory (LSTM) networks for predicting the MR of stabilized base materials with various additives during wet-dry cycles (WDC). A laboratory dataset of 704 records has been used using input parameters, including WDC, ratio of calcium oxide to silica, alumina, and ferric oxide compound, Maximum dry density to the optimal moisture content ratio (DMR), deviator stress (sigma d), and confining stress (sigma 3). The results demonstrate that the LSTM technique is very accurate, with coefficients of determination of 0.995 and 0.980 for the training and testing datasets, respectively. The LSTM model outperforms other developed models, such as support vector regression and least squares approaches, in the literature. A sensitivity analysis study has determined that the DMR parameter is the most significant factor, while the sigma d parameter is the least significant factor in predicting the MR of the stabilized base material under WDC. Furthermore, the SHapley Additive exPlanations approach is employed to elucidate the optimal model and examine the impact of its features on the final result.