Wheat leaf rust, caused by Puccinia triticina Eriks., is recognized as one of the most destructive diseases affecting wheat worldwide, including Iran, resulting in substantial losses in grain yield and quality. This research focused on evaluating the pathogenic factors of nine leaf rust isolates collected from four different climates in Iran, using various differential genotypes. The assessment of leaf rust infection types was conducted on 49 durum and bread wheat genotypes, including susceptible control genotypes and 55 differential genotypes at the seedling stages. The results revealed a significant difference among wheat genotypes in their response to all isolates (p ≤ 0.01). Notably, certain genotypes, such as the Italian landrace (P.S. No4), Shabrang, Chamran2, Mehregan, Shosh, and Gonbad, exhibited resistance to all isolates at the seedling stage, indicating the presence of seedling resistance genes. Additionally, we determined the virulence/avirulence patterns for various resistance genes in the differential genotypes by assessing their responses to different isolates and recording the infection types. The findings indicated that all isolates were virulent on the lines carrying the Lr34 and Lr37 genes, whereas none of the isolates displayed a virulence on the lines carrying the Lr19 gene. This research provides valuable insights into the resistance patterns of wheat genotypes against leaf rust isolates in different climates in Iran, contributing to our understanding of the genetic basis of resistance and aiding in the development of effective strategies for disease management in wheat cultivation. The XGBoost (extreme gradient boosting) algorithm generated the most accurate predictions for the variables thousand grain weight and grain yield, while the MARS (multivariate adaptive regression spline) algorithm generated the most accurate predictions for the variables spike weight, number of grains per spike, and grain weight per spike. For each of these variables, GP (Gaussian process), MARS, and XGBoost achieved the lowest RMSE (root mean square error) values, indicating minimal prediction errors, and the highest R² values, signifying a strong correlation between the predicted and observed data. These prediction performances highlighted the robustness and accuracy of the GP, MARS and XGBoost algorithms in modeling wheat disease severity and its effects on yield outcomes.

Assessing water corrosivity indices is vital for sustainable management, since it damages infrastructure, increase costs, and threaten public health. In this study, the corrosive and scaling behavior of groundwater was modeled and predicted using, Artificial Neural Networks (ANN), Support Vector Machines (SVM), Multivariate Adaptive Regression Splines (MARS), and Random Forest (RF). Three indices were employed: the Langelier Saturation Index (LSI), the Ryznar Stability Index (RSI), and the Puckorius Scaling Index (PSI). The models were developed using 25 years of daily groundwater data from the Dezful-Andimeshk plain in southwestern Iran. In the study area, LSI values ranged from − 8.91 to 0.27, RSI from 8.46 to 18.72, and PSI from − 5.83 to 3.62, indicating that groundwater exhibits both corrosive and scaling tendencies depending on location. SAR (Sodium Adsorption Ratio), pH, and TDS (Total Dissolved Solids) were used as input variables for model development. Among the tested algorithms. The SVM model, performance metrics during the testing phase were as follows: LSI (R² = 0.92, RMSE = 0.11), RSI (R² = 0.81, RMSE = 0.21), and PSI (R² = 0.82, RMSE = 0.21). Overall, model comparisons indicated that all four algorithms achieved acceptable accuracy (R² = 0.80–0.93). However, MARS and ANN consistently provided superior and more stable performance, effectively capturing nonlinear and interactive relationships among the predictors. RF produced competitive results but did not show clear dominance over the other models in this dataset.

Relief wells’ performance gradually decreases over time due to physical and chemical clogging during the operations. Pumping at relief wells is an economic and feasible method to improve the performance of relief wells instead of installing new wells. This paper proposed multi-objective optimization approaches for determining relief well pumping strategy (e.g., pumping rate and the number of pumping wells). Firstly, a three-dimensional transient seepage of Yangtze River levee with relief wells is simulated using MODFLOW model. The well pumping strategy is optimized by minimizing the average safety factor deficit to a defined threshold at the cross-section of relief wells, minimizing total pumping rate and minimizing the number of pumping wells. Then, non-dominated sorting genetic algorithm-II (NSGA-II) is used to derive the alternative optimal pumping strategies. To reduce enormous computational burden within the multi-objective optimization, a nonparametric regression procedure, i.e., multivariate adaptive regression splines (MARS), is used to establish an intelligent surrogate model for evaluating the safety factor of levee with relief wells instead of repetitive MODFLOW simulations. Finally, the best pumping strategy among Pareto solutions is selected by entropy weight and the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS). The proposed approaches are tested on the Yangxin Yangtze River Dike in Hubei Province. Results show that the proposed approaches can objectively determine the final optimal relief well pumping operation by balancing the safety of levee and economic efficiency. Compared with the traditional method, the proposed approaches only require pumping groundwater from two wells for the study area, where the number of pumping wells is reduced by 50%, and the total pumping rate is decreased by 37.5%. Moreover, the proposed approaches significantly reduce the computation time from several thousand hours of repetitive numerical simulations to just one minute, providing a feasible tool for relief well maintenance.

The widespread adoption of bulk metallic glasses (BMGs) in aerospace and biomedical industries requires topology-optimized architectures that conventional manufacturing cannot achieve. In response, BMGs have been investigated for powder bed fusion (PBF), but the process remains challenging due to narrow thermal windows, expensive feedstock, and limited data. This study introduces a constrained multi-objective Bayesian optimization framework to optimize key PBF printing parameters, including laser power and scan speed, to maximize hardness while preserving the amorphous state of the printed BMG. Hardness is optimized as the primary objective with density incorporated in the scalarization to regularize the search space, and amorphous retention is enforced through a feasibility probability learned by a logistic classifier. Surrogate models are compared, including Gaussian process, Bayesian additive regression trees, Bayesian multivariate adaptive regression splines (BMARS), and a Bayesian attention neural network. Acquisition scores are computed with constrained expected improvement and are maximized on a uniform grid over power and velocity. Superior predictive accuracy is obtained with BMARS, and 95% credible intervals are calibrated to the measurements. A high-hardness region at high power and low velocity is localized by the surrogates. A fully amorphous sample at 60 W and 1300 mm/s is produced, and a hardness of 1010.4 HV is measured in agreement with the predicted high-hardness band. In conclusion, the study establishes a data-efficient process-window discovery method for BMG PBF, produces an interpretable process map, and demonstrates a screening framework suitable for constrained experimental budgets.

ABSTRACT Accurate prediction of river water quality parameters is essential for environmental protection and sustainable agricultural resource management. This study presents a novel framework for estimating potential salinity in river water in arid and semi‐arid regions by integrating a kernel extreme learning machine (KELM) with a boosted salp swarm algorithm based on differential evolution (KELM‐BSSADE). A dataset of 336 samples, including bicarbonate, calcium, pH, total dissolved solids and sodium adsorption ratio, was collected from the Idenak station in Iran and was used for the modelling. Results demonstrated that KELM‐BSSADE outperformed models such as deep random vector functional link (dRVFL), general regression neural network (GRNN), multivariate adaptive regression spline (MARS), online sequential extreme learning machine (OSELM) and extreme gradient boosting decision tree (XGBoost) when compared with observed river salinity data. Also, the KELM‐BSSADE model effectively identified optimal inputs through the Boruta‐XGBoost (B‐XGB) feature selection method. Four metaheuristic‐based KELM models were developed, utilizing grey wolf optimizer, whale optimization, slime mould algorithm and equilibrium optimizer, further illustrating the capability of KELM‐BSSADE in estimating potential salinity in river water. By accurately estimating potential salinity, KELM‐BSSADE can assist in optimizing irrigation practices, ensuring that agricultural demands are met while minimizing the risk of salinity‐related crop damage.

Introduction/Background: The Martin-Hopkins equation is widely validated and used to estimate low-density lipoprotein cholesterol (LDL-C). However, some laboratories find implementation challenging and a machine learning-based approach may help overcome this by simplifying implementation to a single equation. Research Question/Hypothesis: This study aims to develop a machine-learning based equation for estimating LDL-C using multivariate adaptive regression spline (MARS) and compare its performance to the Friedewald (LDL-C F ), Sampson-NIH (LDL-C S ), and original Martin-Hopkins (LDL-C MH ) equations. Methods/Approach: We used data from the Very Large Database of Lipids (VLDbL), which includes lipid measurements collected using VAP ultracentrifugation between October 1, 2015, and June 30, 2019. The study included 4,939,528 patients with complete lipid panel data, who were randomly assigned to a training set (n=3,292,889) and a test set (n=1,646,639). A MARS model was developed to estimate LDL-C ( Figure 1 ). The accuracy of LDL-C estimation was assessed using bias, root mean squared error (RMSE), and concordance with guideline-based categories between estimated and ultracentrifugation-measured LDL-C values. Results/Data: The new equation (LDL-C MH-MARS ) showed a very low median bias (-0.14 mg/dL, IQR: -2.11–1.78), comparable to LDL-C MH (0.26 mg/dL, IQR: -1.62–2.37). The median bias was higher for LDL-C S (1.72 mg/dL, IQR: -1.29–4.14) and LDL-C F (-0.20 mg/dL, IQR: -4.40–2.60). The median difference between the MARS and original Martin-Hopkins equations was -0.51 mg/dL (IQR: -1.24–0.04), further suggesting comparability between the two methods. RMSE was lowest for LDL-C MH-MARS (4.68) and LDL-C MH (4.90), followed by LDL-C S (5.79) and LDL-C F (7.25). The proportion of patients correctly classified to clinical categories was nearly identical for LDL-C MH-MARS (89.69%) and LDL-C MH (89.60%), but lower for LDL-C S (86.28%) and LDL-C F (83.10%) ( Figure 2 ). LDL-C S and LDL-C F underestimated LDL-C in 39% and 60% when classifying LDL-C <70 mg/dL in patients with triglyceride concentrations of 200–399 mg/dL, whereas this improved to <18% with both LDL-C MH-MARS and LDL-C MH ( Figure 3 ). Conclusions: The new machine learning-based LDL-C equation provides comparable results to the original Martin-Hopkins method while simplifying implementation to a single equation. The implementation of LDL-C MH-MARS would be straightforward with a single line of code in laboratory information systems.

ABSTRACT Stochastic reservoir characterisation relies on the careful integration of geological modelling and geophysical data, enabling prediction and uncertainty quantification for reservoir decision‐making. In this paper, we address some of the computational challenges associated with Bayesian reservoir characterisation, focusing on key obstacles: demanding geophysical forward modelling, high dimensionality of spatial variables and effective posterior sampling of reservoir variables given geophysical data and well information. Leveraging a pseudo‐Bayesian approach, we replace the intricate forward model for seismic amplitude‐versus‐offset data with a computationally efficient multivariate adaptive regression splines method, resulting in a 34‐times acceleration in computations. For handling high‐dimensional variables modelled by Gaussian random fields, we employ a fast Fourier transform technique. We use a preconditioned Crank–Nicolson method for efficient Markov chain Monte Carlo sampling from the posterior of the reservoir variables. The approach is motivated by challenging reservoir conditions at the Alvheim field in the North Sea, where we demonstrate our approach for Bayesian posterior sampling of oil and gas saturation and clay content conditional on seismic amplitude data and well information. We compare our results with those obtained from an approximate ensemble‐based Kalman method for posterior sampling.

This study examines the potential of Multivariate Adaptive Regression Splines (MARS) in predicting recorded heights above mean sea level within the Tarkwa Local Geodetic Reference Network in Ghana. Logistical and computational constraints of conventional techniques, such as spirit levelling and geostatistical interpolation, drive the assessment of MARS as a strong soft computing substitute. The MARS model was trained and verified using field-measured data gathered using a Total Station DTM 122A, and its performance was compared against the Polynomial Regression model (PRM) and Kriging models. Each model technique was assessed based on statistical models such as arithmetic mean absolute error (AMAE), arithmetic mean squared error (AMSE), arithmetic root mean squared error (ARMSE), arithmetic standard deviation (ASD), correlation coefficient (R), and coefficient of determination (R2). Statistical measures showed MARS's better accuracy utilizing near-perfect correlation (AMAE: 1.7963E-06 m; AMSE: 8.6775E-12 m) and low error margins. The results show MARS to be a possible, high-precision solution for orthometric height calculation, hence improving Ghana's geodetic network uses in environmental management, building, and surveying. This work not only confirms the effectiveness of MARS but also provides a basis for improving height measurement methods in local geodetic systems.

The study focuses on the experimental investigation of the impact of using coconut oil (CO) as a phase-change material (PCM) for heat storage on the root-zone temperature within a greenhouse in Adana, Türkiye. The study examines the efficacy of PCM as latent heat-storage material and predicts root-zone temperature using three machine learning algorithms. The dataset used in the analysis consists of 2658 data at hourly resolution with six variables from February to April in 2022. A greenhouse with PCM shows a remarkable increase in both ambient (0.9–4.1 °C) and root-zone temperatures (1.1–1.6 °C) especially during the periods without sunlight compared to a conventional greenhouse. Machine learning algorithms used in this study include Multivariate Adaptive Regression Splines (MARS), Support Vector Regression (SVR), and Extreme Gradient Boosting (XGBoost). Hyperparameter tuning was performed for all three models to control model complexity, flexibility, learning rate, and regularization level, thereby preventing overfitting and underfitting. Among these algorithms, R2 values for testing data listed from largest to smallest are MARS (0.95), SVR (0.96), and XGBoost (0.97), respectively. The results emphasize the potential of machine learning approaches for applying thermal energy storage systems to agricultural greenhouses. In addition, it provides insight into a net-zero energy greenhouse approach by storing heat in a bio-based PCM, alongside its implementation and operational procedures.

ObjectiveMild cognitive impairment (MCI) signals cognitive decline beyond normal aging and increases dementia risk. Early identification enables preventative interventions, yet many patients in primary care go undetected. This study examines whether machine learning (ML) models can predict MCI using routinely collected personal, social, and functional lifestyle factors and identifies the most important predictors.MethodsData from round 2 and 3 of the National Social Life, Health, and Aging Project was used, including 4,586 older adults with complete Montreal Cognitive Assessment (MoCA) scores. Predictors included demographics, childhood experiences, health behaviors, psychosocial measures, and functional difficulties. Eight ML models—including elastic net, multivariate adaptive regression splines, random forest, oblique random forest, boosted trees, decision trees, and a stacked ensemble—were trained and tuned using repeated cross-validation, with 20% of the dataset withheld for final testing. Model performance was assessed using area under the receiver operator curve (AUROC), accuracy, sensitivity, specificity, and Matthew’s correlation coefficient (MCC).ResultsMost models achieved good discrimination (AUROC > 0.8), with the stacked ensemble performing best (AUROC = 0.823; MCC = 0.462). The best individual model was logistic regression (AUROC = 0.818). Across models, key predictors of MCI included age, ethnicity, functional difficulties, social disconnectedness, and perceived stress.DiscussionLogistic regression outperformed more complex machine learning models, providing the best combination of predictive accuracy and interpretability for identifying MCI. Across models, age, ethnicity, functional difficulties, social disconnectedness, and stress consistently emerged as key predictors, highlighting their central role in cognitive health. These findings suggest that psychosocial and functional measures can serve as practical indicators for those who need to be screened early for MCI, offering an opportunity for timely intervention and support. However, future work should include longitudinal data and clinical diagnoses to validate and refine these predictive tools.

The substantial energy consumption and associated CO2 emissions from industrial operations pose significant environmental and economic challenges for factories and surrounding communities. Within the context of industrial energy management, the steel industry represents a major energy consumer. The imperative to optimize energy use in this sector is driven by a combination of environmental concerns, economic incentives, and technological advancements. This study presents a machine learning model that integrates the whale optimization algorithm (WOA) with multivariate adaptive regression splines (MARS) to forecast electric energy consumption. Utilizing a dataset comprising 35,040 real-world energy consumption records from Gwangyang Steelworks in South Korea, the model was benchmarked against other regression techniques (ridge, lasso, and elastic-net), demonstrating that the proposed WOA-MARS approach achieves a significant improvement in the RMSE (vs. elastic-net or lasso regression techniques) while maintaining interpretability through hinge function analysis. The WOA-tuned MARS model achieves a coefficient of determination (R2) of 0.9972, underscoring its effectiveness for energy optimization in steel manufacturing. The key findings reveal that CO2 emissions and reactive power variables are the strongest predictors.

PurposeTo develop and compare multiple machine learning (ML) algorithms with traditional logistic regression for predicting in-hospital cardiac arrest (IHCA) using comprehensive electronic health record data, with the goal of improving early risk stratification beyond conventional early-warning scores and providing potential integration into hospital early warning systems for timely clinical intervention.Patients and MethodsWe performed a retrospective case-control study at a large tertiary medical center, including 800 IHCA cases and 3,464 controls. Candidate predictors comprised demographics, comorbidities, vital signs, and laboratory measurements. Five models-logistic regression, decision tree, random forest, XGBoost, and multivariate adaptive regression splines (MARS)-were trained and validated. Performance was evaluated using area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, specificity, and F1 score.ResultsXGBoost yielded strong discrimination and the highest accuracy (AUC 0.909; accuracy 0.883), while random forest showed comparable discrimination (AUC 0.910) with slightly lower accuracy (0.876). Logistic regression performed robustly but lower than ML models (AUC 0.895; accuracy 0.876). ML models consistently identified clinically meaningful predictors-including blood urea nitrogen, heart rate, and pre-existing heart failure-offering insights beyond traditional regression.ConclusionIntegrating ML approaches with conventional regression enhances IHCA risk prediction by capturing non-linear relationships and interactions while retaining the interpretability of regression. These approaches could strengthen hospital early-warning systems, enabling earlier detection and intervention, and ultimately improving patient outcomes.

Background: Uric acid (UA) is linked to gout, renal dysfunction, and cardiovascular disease. Prior studies often assume linear relationships, potentially oversimplifying physiological complexity. Methods: We analyzed data from 5200 healthy Taiwanese men. Demographic, biochemical, lifestyle, and inflammatory variables were assessed using Pearson correlation, multiple linear regression (MLR), and multivariate adaptive regression splines (MARS), an interpretable machine learning method for detecting nonlinear, threshold-based effects. Results: Pearson correlation showed broad linear associations, whereas MARS identified fewer but more physiologically meaningful predictors. Waist-to-hip ratio (WHR) had a strong threshold effect, influencing UA only below 0.969. Creatinine showed a nonlinear impact, becoming substantial above 0.97 mg/dL, suggesting a renal threshold within the “normal” range. Calcium and high-sensitivity C-reactive protein (hs-CRP) each displayed inflection points (9.5 mg/dL and 3.38 mg/L, respectively), indicating range-specific effects. Notably, betel nut exposure, nonsignificant in linear models, emerged in MARS as a predictor with a complex, non-binary association with UA metabolism. Predictive performance was comparable (RMSE: 1.6694 for MARS vs. 1.6666 for MLR), but MARS offered superior interpretability by highlighting localized nonlinear effects. Conclusions: MARS modeling revealed critical nonlinear, threshold-dependent associations between UA and WHR, creatinine, calcium, hs-CRP, and betel nut exposure, which were not captured by conventional methods. These findings underscore the value of interpretable machine learning in metabolic research and suggest precise thresholds for clinical risk stratification.

Unstable soil-related disasters are a concern for modern urbanization and development. Hence, it is necessary to stabilize the soil and estimate the geotechnical properties using advanced methodologies, one of the best-suited methods for safe ground to construct any infrastructure. Accordingly, this study has provided machine learning techniques, namely Multiple Linear Regression (MLR), Support Vector Regression (SVR), Multivariate Adaptive Regression Splines (MARS), and Random Forest (RF), to estimate the CBRs of soil. The dataset used in the study comprised 15 observations. The data consists of input variables including the percentages of fly ash (%FA), cement (%C), and coir fiber (%CF), along with optimum moisture content (OMC), maximum dry density (MDD), and soaked California Bearing Ratio (CBR). The dataset is divided into a training set (55% of the total data) and a testing set (45%). The performance of the developed models was achieved using root mean square error (RMSE), mean absolute error (MAE), correlation coefficient (r), and volumetric efficiency (VE). For the training set, MLR achieved RMSE = 1.34, MAE = 1.08, r = 1.00, and VE = 0.96, while MARS recorded 7.06, 6.03, 0.92, and 0.77, respectively. In the testing set, MLR again outperformed with 0.86, 0.80, 1.00, and 0.98, compared to MARS, which obtained 3.22, 2.59, 0.99, and 0.92. This study introduces a novel approach by evaluating the predictive capabilities of linear and non-linear machine learning algorithms for soaked CBR estimation using a limited geotechnical dataset, thereby providing insights into model robustness and reliability in data-constrained scenarios.

Simple SummaryRaptors, as apex predators, serve as valuable bioindicators for assessing the impacts of climate change because of their specialized ecological traits, which render them particularly susceptible to environmental alterations. Globally, raptors are experiencing significant conservation concerns, with approximately 52% of species exhibiting declining populations and 18% being classified as threatened. Despite this, the effect of climate change on raptors is poorly studied in the Eastern Himalayan region. Three species, Falco severus, Gyps tenuirostris and Haliaeetus leucoryphus, were selected based on their conservation status in the region. This study provides a comprehensive assessment of climate change impacts on raptors in the northeastern part of the Eastern Himalayas, utilizing ensemble species distribution modeling for the projected periods 2041–2060 and 2061–2080. The future projections indicate a substantial decline in suitable habitats: Falco severus is projected to lose 33–41%, Gyps tenuirostris may lose 53–96%, and Haliaeetus leucoryphus is anticipated to experience a loss of approximately 94–99% of its suitable habitats.Raptors, being at top of the food chain, serve as important models to study the impact of changing climate, as they are more vulnerable due to their unique ecology. They are vulnerable to extinction, with 52% species declining population and 18% are threatened globally. The effect of climate change on raptors is poorly studied in the Eastern Himalayan region. The present study offers a complete investigation of climate change effects on the raptors in the northeast region of the Eastern Himalayas, employing ensemble species distribution modeling. The future predictions were employed to model the climate change across two socioeconomic pathways (SSP) i.e. SSP245 and SSP585 for the periods 2041–2060 and 2061–2080. Specifically, five algorithms were employed for the ensemble model, viz. boosted regression tree (BRT), generalized linear model (GLM), multivariate adaptive regression splines (MARS), maximum entropy (MaxEnt) and random forest (RF). The study highlights worrying results, as only 10.5% area of the NE region is presently suitable for Falco severus, 11.4% for the critically endangered Gyps tenuirostris, and a mere 6.9% area is presently suitable for the endangered Haliaeetus leucoryphus. The most influential covariates were precipitation of the driest quarter, precipitation of the wettest month, and temperature seasonality. Future projection revealed reduction of 33–41% in suitable habitats for F. severus, G. tenuirostris is expected to lose 53–96% of its suitable habitats, and H. leucoryphus has lost nearly 94–99% of its suitable habitats. Such decline indicates apparent habitat fragmentation, with shrinking habitat patches.

Comparative analysis of Biomod2 statistical and machine learning methods for Lactarius deliciosus distribution in Refahiye, Turkiye.

Abstract The evaluation of public policies is an increasingly important issue that demands greater attention. This study examines the impact of the Severino Law (2012) on corruption control and inequality in Italy. Using a robust quantitative approach for the period 2002–2022, we apply difference-in-differences (DiD), the synthetic control method (SCM), and a multivariate adaptive regression splines (MARS) model. Our findings indicate that while the law significantly improved corruption control—albeit with a delayed effect—it did not reduce inequality as anticipated, which instead continued to rise. This outcome stands in contrast to the predictions of much of the empirical literature. The results suggest that the direct effects of corruption control on inequality can be highly idiosyncratic and do not operate in a linear or isolated manner. Rather, they depend on the extent to which corrupt behaviors are reduced and on the presence of complementary policies beyond anti-corruption reforms to effectively address structural inequality.

In the field of environmental sustainability, the preservation of water resources and the maintenance of water quality are of utmost importance. The aim of this study is to develop a predictive model for assessing the water quality of the Kızılırmak river by integrating Principal Component Analysis (PCA) and Multivariate Adaptive Regression Splines (MARS) methodologies. To assess water quality, surface water samples obtained from six distinct locations during the 2022-2023 period were analyzed with respect to seventeen physicochemical parameters. The first stage of the present study was the determination of the most informative variables in the water quality data set using the dimensionality reduction method PCA. In the second phase, a predictive model was developed using the MARS algorithm based on the principal components derived from the PCA-reduced dataset. The MARS algorithm was proposed to predict Water Quality Index (WQI) values using this reduced dataset. A coefficient of determination (R2) value of 0.997 was achieved for predicting the WQI in the study area. According to the results of this study, the MARS model developed using PCA demonstrated high precision and performance in estimating the WQI. This methodological framework clarified the interactions between parameters in water quality assessment studies, allowing for a comprehensive analysis of their overall effects on WQI.

Background: Myocardial perfusion scintigraphy (MPS) is an important tool for evaluating ischemia in diabetic populations. However, applications of advanced predictive models like multivariate adaptive regression splines (MARS) to estimate summed stress scores (SSS) are lacking. Methods: In this study, 1028 diabetic women undergoing Thallium-201 MPS were analyzed. The dataset was split into training (80%) and testing (20%) subsets. MARS and multiple linear regression (MLR) models were constructed to predict SSS, and their performance was evaluated using root mean square error (RMSE), relative absolute error (RAE), root relative squared error (RRSE), Mean Absolute Percentage Error (MAPE), and Symmetric Mean Absolute Percentage Error (SMAPE). Results: On the testing dataset, the MARS model outperformed the MLR model across all metrics, with an RMSE of 3.25 compared to 3.89 for MLR, an RAE of 0.52 vs. 0.64, and an RRSE of 0.53 vs. 0.67. Similar trends were observed in MAPE (18.7% vs. 22.1%) and SMAPE (17.3% vs. 20.5%). Conclusions: The superior predictive accuracy of the MARS model suggests its potential to enhance non-invasive myocardial risk stratification in diabetic women.

Developing universal hydrological models for modeling urban catchments remains one of the major challenges in contemporary hydrology. This study aimed to create a model that integrates catchment characteristics, sewer network topology, sewer storage capacity, and rainfall data, along with a sensitivity analysis of input parameters. The goal was to evaluate the potential of advanced analytical methods, specifically, Multivariate Adaptive Regression Splines (MARS) and soft-sensor technology, to improve peak flow (Qm) forecasting in stormwater systems. The results showed that combining MARS models with soft sensors yields high forecasting accuracy (R² = 0.96, RMSE = 0.038), even under variable rainfall conditions. However, the development of universally applicable model relationships proved challenging due to difficulties in parameterizing the model under changing rainfall scenarios. Additionally, the inclusion of a risk analysis method also enabled consideration of sewer network capacity and introduced a safety margin coefficient to assess system flexibility under future climate conditions. While the proposed approach does not lead to the creation of universal tools, it offers valuable insights for further research on adapting sewer systems to evolving hydrological conditions. The findings suggest promising directions for the development of cost-free, zero-emission soft sensors and models adaptable across diverse urban catchments.