The compressive strength (Fc) of cement mortar (CM) is a key parameter in ensuring the mechanical reliability and durability of cement-based materials. Traditional testing methods are labor-intensive, time-consuming, and often lack predictive flexibility. With the increasing adoption of machine learning (ML) in civil engineering, data-driven approaches offer a rapid, cost-effective alternative for forecasting material properties. This study investigates a wide range of supervised linear and nonlinear ML regression models to predict the Fc of CM. The evaluated models include linear regression, ridge regression, lasso regression, decision trees, random forests, gradient boosting, k-nearest neighbors (KNN), and twelve neural network (NN) architectures, developed by combining different optimizers (L-BFGS, Adam, and SGD) with activation functions (tanh, relu, logistic, and identity). Model performance was assessed using the root mean squared error (RMSE), coefficient of determination (R²), and mean absolute error (MAE). Among all models, NN_tanh_lbfgs achieved the best results, with an almost perfect fit in training (R² = 0.9999, RMSE = 0.0083, MAE = 0.0063) and excellent generalization in testing (R² = 0.9946, RMSE = 1.5032, MAE = 1.2545). NN_logistic_lbfgs, gradient boosting, and NN_relu_lbfgs also exhibited high predictive accuracy and robustness. The SHAP analysis revealed that curing age and nano silica/cement ratio (NS/C) positively influence Fc, while porosity has the strongest negative impact. The main novelty of this study lies in the systematic tuning of neural networks via distinct optimizer–activation combinations, and the integration of SHAP for interpretability—bridging the gap between predictive performance and explainability in cementitious materials research. These results confirm the NN_tanh_lbfgs as a highly reliable model for estimating Fc in CM, offering a robust, interpretable, and scalable solution for data-driven strength prediction.

This paper presents a complete model for thermoelectric modules (TEMs) in order to be used in circuit simulations in MATLAB/Simulink software. The purpose of this study is to develop a single block like other circuit elements in the Simulink library by combining both thermoelectric generator (TEG) and thermoelectric cooler (TEC) module parameters and their operations. Since TEG and TEC are both thermoelectric devices, mostly whose main material, such as Bi₂Te₃, is the same for similar temperature ranges, a TEG can be operated as TEC and vice versa a TEC can be operated as TEG. In some systems, such as PV-thermoelectric (TE), fuel cell-TE, and battery temperature management of electric vehicles, the modules can be used in both modes. Also, in some cases of these systems, sometimes the TEM may need to be operated as a generator, and sometimes as a cooler. The bidirectional operation of the TEM is a cost-saving approach since the 2 types of TEMs are not needed to be installed together. This paper proposes a novel model whereas the studies up to date in the literature have not been fully compatible with the bidirectional usage of the modules. The proposed model deals with the polarities of both the thermal and the electrical ports of the TEM in order to decide the mode of operation, that is TEG or TEC. Maximum parameters and performance parameters for both TEG and TEC operations are included in the module block so as to observe and control the device for desired operating conditions. It is seen that the simulations with the temperature-dependent parameters have given very close results to the graphics of the datasheet. However, simulations in which the Thomson heat isn’t ignored have more deviations compared to the datasheet. Moreover, the simulations with the constant material properties have given better results than the Thomson heat included ones in the cases of the higher ΔT levels. While ΔT is 10 °K, COP_max without the Thomson heat has been found as 3.399 which is the same as the datasheet. Meanwhile, COP_max with the Thomson heat has been found as 3.7147, and COP_max with the constant properties has been found as 3.8057. However; while ΔT is 50 °K, COP_max is 0.247 without the Thomson heat which is the same as the datasheet, but COP_max is 0.341 with the Thomson heat, and COP_max is 0.268 with constant material properties. In addition, it is shown that, under variable polarities of thermal and electrical ports, bidirectional operation of the TEM is successfully achieved and the TEM is operated as both TEG and TEC in the same simulation.

The affordability, ease of manufacturing, and assembly efficiency of cold-formed steel profiles have contributed to their widespread use in structural applications. However, the presence of holes in these profile webs is likely to reduce their mechanical resistance. This study explores the bending behavior of a built-up box section constructed using lipped and unlipped C-profiles, which are commonly utilized in the construction industry. The investigation focuses on the influence of self-drilling screw layout density and hole distribution within the section. A total of 30 different models were analyzed, considering three primary variables: the spacing of self-drilling screws, hole diameter, and the number of holes. The steel profiles were connected using self-drilling screws with spacing intervals of 100, 200, and 400 mm. Key parameters, such as moment capacity, effects on elastic zones, shear forces on screws, and ductility, were examined in relation to these variables. The findings indicate that reducing screw spacing and increasing the number of holes are crucial design factors for improving joint strength. However, while greater screw spacing enhances ductility, it leads to lower plastic deformation rates. Additionally, optimizing the number of holes in the section proved to be an effective strategy for improving ductility in the analyzed models. Mathematical evaluation confirmed that hole number and screw spacing significantly affect moment capacity and estimation stability, highlighting the need for their joint optimization in structural design.

In this study, in order to reduce the electricity consumption / cost of the Karaman Wastewater Treatment plant, which has the highest flow and electricity consumption in Sakarya, the arrangements that can be made during the working hours of the treatment plant and the generation of electricity by the MHP that can be installed between the outlet of the facility and the discharge point of the wastewater are examined. A change of 17% in the electricity bill costs and up to 50% in the flow rates were observed. Wastewater treatment plant projects keep their capacities high, considering the needs of the next 20-30 years. Based on these results, it is determined that if the electricity consumption amounts in the peak time interval where the unit price of electricity is high is shifted to day and night time zones, the electricity costs of the facilities will decrease. In addition, it has been shown that a part of its own electricity consumption will be met by the micro hydro power plant to be installed between the outlet of the facility and the discharge point.

Introduction: This study aims to predict mortality using chest radiographs obtained on the first day of intensive care admission, thereby contributing to better planning of doctors’ treatment strategies and more efficient use of limited resources through early and accurate predictions. Methods: We retrospectively analyzed 510 ICU patients. After data augmentation, a total of 3019 chest radiographs were used for model training and validation, while an independent, non-augmented test set of 100 patients (100 images) was reserved for final evaluation. Seventy-four (74) radiomic features were extracted from the images and analyzed using machine learning algorithms. Model performances were evaluated using the area under the ROC curve (AUC), sensitivity, and specificity metrics. Results: A total of 3019 data samples were included in the study. Through feature selection methods, the initial 74 features were gradually reduced to 10. The Subspace KNN algorithm demonstrated the highest prediction accuracy, achieving AUC 0.88, sensitivity 0.80, and specificity 0.87. Conclusions: Machine learning algorithms such as Subspace KNN and features obtained from PAAC radiographs, such as GLCM Contrast, Kurtosis, Cobb angle, Haralick, Bilateral Infiltrates, Cardiomegaly, Skewness, Unilateral Effusion, Median Intensity, and Intensity Range, are promising tools for mortality prediction in patients hospitalized in the internal medicine intensive care unit. These tools can be integrated into clinical decision support systems to provide benefits in patient management.

Aim: In this study, it has been proposed to solve the Optimal Reactive Power Dispatch (ORPD) problem with Genetic Algorithm (GA) in order to minimize the active power losses and voltage fluctuations in the 114 bus Northwestern Anatolian (NWA) power system. Method: The ORPD problem, which is modeled as a nonlinear optimization problem, has been solved by using GA. The ORPD problem has been applied to a large 114 bus power system in Northwestern Anatolia. By solving the ORPD problem, terminal voltages of generators and reactive power values of shunt compensators, which are optimal control variables, have been found. Control variables values to minimize active power losses and voltage fluctuations have been obtained by using GA. Results: The proposed approach has successfully solved the ORPD problem by finding the voltage values of the generators and the sizes of the reactive power sources which are the control variables of the optimization problem. The aim of the ORPD problem is to minimize power losses and to reduce the voltage deviations in the system considering some operational constraints. The proposed approach has been applied to the 114 bus NWA power system and it has been observed that the active power losses are reduced by 4.63% and the bus voltage deviations are decreased from 0.08 to 0.0798 after ORPD. Conclusion: By solving the ORPD problem, active power losses and voltage fluctuations at buses are simultaneously minimized by ensuring that all bus voltages remain within the defined limits in the 114-bus NWA power system. Simulation studies have shown that with the proposed ORPD approach, the bus voltages remain within the specified limits and the active power losses are reduced compared to the first case. The reduction in bus voltage deviations means that the system becomes more stable in terms of stability. Thus, while saving is achieved by increasing the efficiency in the power system, it is also possible to monitor the stability of the system.