Accurate prediction of the uniaxial compressive strength (UCS) of concrete is crucial for ensuring the safety, durability, and performance of structures in construction. This study presents a predictive model using a multilayer perceptron (MLP), to estimate UCS based on key input parameters such as water-cement ratio, aggregate size, curing time, water and cement content. The MLP model was trained and validated using a dataset comprising 120 cubic laboratory-tested concrete samples (15cm × 15cm × 15cm) with varying compositions for normal construction materials. Performance of the model was evaluated using statistical metrics (split into training and testing sets as 70%-30%), showing that the MLP-based approach provides accurate and reliable predictions compared to traditional regression models. The proposed method offers a practical, efficient tool for geotechnical engineers to assess concrete strength, potentially reducing the need for extensive experimental testing and enhancing quality control in concrete production.

The uniaxial compressive strength (UCS) of intact rocks is one of the key parameters in the course of site characterizations. The isotropy/anisotropy condition of the UCS of intact rocks is dependent on the internal structure of the rocks. The rocks with a random grain structure exhibit an isotropic behavior. However, the rocks with a linear/planar grain structure generally behave transversely-isotropic. In the latter case, the UCS of intact rocks must be determined by a set of laboratory tests on the oriented rock samples. There are some empirical relations available to describe the strength of these rocks. Though characterization of transversely-isotropic rocks is practically a 3D problem, but these relations provide only a 2D description. In this paper, a method is proposed to provide a 3D description of UCS of transversely-isotropic rocks. By means of this formulation, one can determine UCS along with any arbitrary spatial direction. Also, a representative illustration of UCS is proposed in the form of contour-plots on a lower hemisphere Stereonet. The method is applied to an actual case study from the Kanigoizhan dam site located in the Kurdistan Province (Iran). Application of the proposed method to the phyllite rocks of this site show that the direction perpendicular to the dam axis exhibits the most anisotropic behavior. Hence, it is essential to take the strength anisotropy into account during the relevant analysis. The results obtained, together with the statistical variation of UCS, provide a practical approach to select the proper values of UCS according to the scope of the analysis.

In this study, several soft computing analyses are performed to build some predictive models to estimate the uniaxial compressive strength (UCS) of the pyroclastic rocks from central Anatolia, Turkey. For this purpose, a series of laboratory studies are conducted to reveal physico-mechanical rock properties such as dry density (ρd), effective porosity (ne), pulse wave velocity (Vp), and UCS. In soft computing analyses, ρd, ne, and Vp are adopted as the input parameters since they are practical and cost-effective non-destructive rock properties. As a result of the soft computing analyses based on the classification and regression trees (CART), multiple adaptive regression spline (MARS), adaptive neuro-fuzzy inference system (ANFIS), artificial neural networks (ANN), and gene expression programming (GEP), five robust predictive models are proposed in this study. The performance of the proposed predictive models is evaluated by some statistical indicators, and it is found that the correlation of determination (R2) value for the models varies between 0.82 – 0.88. Based on these statistical indicators, the proposed predictive models can be reliably used to estimate the UCS of the pyroclastic rocks.

The uniaxial compressive strength of weak rocks (UCSWR) is among the essential parameters involved for the design of underground excavations, surface and underground mines, foundations in/on rock masses, and oil wells as an input factor of some analytical and empirical methods such as RMR and RMI. The direct standard approaches are difficult, expensive, and time-consuming, especially with highly fractured, highly porous, weak, and homogeneous rocks. Numerous endeavors have been made to develop indirect approaches of predicting UCSWR. In this research work, a new intelligence method, namely relevance vector regression (RVR), improved by the cuckoo search (CS) and harmony search (HS) algorithms is introduced to forecast UCSWR. The HS and CS algorithms are combined with RVR to determine the optimal values for the RVR controlling factors. The optimized models (RVR-HS and RVR-CS) are employed to the available data given in the open-source literature. In these models, the bulk density, Brazilian tensile strength test, point load index test, and ultrasonic test are used as the inputs, while UCSWR is the output parameter. The performances of the suggested predictive models are tested according to two performance indices, i. e. mean square error and determination coefficient. The results obtained show that RVR optimized by the HS model can be successfully utilized for estimation of UCSWR with R2 = 0. 9903 and MSE = 0. 0031203.

This study utilizes a support vector machine (SVM) supervised learning algorithm to predict the uniaxial compressive strength (UCS) of lightweight concrete (LWC). Implemented in Python 3, the model leverages a dataset of 120 samples of LWC, which was divided into training and testing sets in a 70%-30% split. Key performance indicators, including mean absolute error (MAE), mean squared error (MSE), and root mean squared error (RMSE), were assessed for both sets. A confusion matrix was also generated to evaluate the classification accuracy. The predictive modeling results show that the SVM algorithm achieved an accuracy of 83. 63% and a precision of 85. 11% in predicting the UCS of LWC samples. The calculated error metrics were promising, with MAE, MSE, and RMSE values at 0. 346, 0. 329, and 0. 331, respectively. These findings suggest that the SVM model is capable of accurately predicting UCS for lightweight concrete, demonstrating its potential for application in concrete strength assessment. As results, the SVM model offers an efficient predictive approach for estimating UCS in lightweight concrete, providing valuable insights for material testing and construction planning. This work highlights the utility of machine learning in advancing accuracy and reliability in civil engineering materials science.