Checking for leakage flow in hydraulic and marine structures during design practice is a crucial step, as uncontrolled leakage can cause irreparable damage. . Soft computing methods can be used to easily model, analyze and control complex systems. This study uses Support Vector Machine (SVM) method to predict leakage discharge of coastal dykes. Five different models are used to achieve this goal, with parameters including the length of the cutoff blanket, dyke depth, and water head considered. The best Support Vector Machine model is checked using a multivariate adaptive Regression spline model (MARS) for prediction. Results show that the model including all parameters predicts settlement discharge with very good accuracy compared to the laboratory model, with a coefficient of determination and root mean square coefficient of 0. 949 and 0. 058 respectively in the test stage and 0. 93 and 0. 06 in the test phase estimates. The dyke depth parameter has the greatest effect on leakage flow, while the water head has the least effect among input parameters to the model. Although the adaptive Regression multivariate spline model accurately estimates the annual dyke leakage flow rate, it is less accurate than the Support Vector Machine method.

Background: Diagnosing of obstructive sleep apnea (OSA) is an important subject in medicine. This study aimed to compare the performance of two data mining techniques, Support Vector Machine (SVM), and logistic Regression (LR), in diagnosing OSA. The best‑ fit model was used as a substitute for polysomnography (PSG), which is the gold standard for diagnosing this disease. Materials and Methods: A total of 250 patients with sleep problems complaints and whose disease had been diagnosed by PSG and referred to the Sleep Disorders Research Center of Farabi Hospital, Kermanshah, between 2012 and 2015 were recruited in this study. To fit the best LR model, a model was first fitted with all variables and then compared with a model made from the significant variables using Akaike’ s information criterion (AIC). The SVM model and radial basis function (RBF) kernel, whose parameters had been optimized by genetic algorithm, were used to diagnose OSA. Results: Based on AIC, the best LR model obtained from this study was a model fitted with all variables. The performance of final LR model was compared with SVM model, revealing the accuracy 0. 797 versus 0. 729, sensitivity 0. 714 versus 0. 777, and specificity 0. 847 vs. 0. 702, respectively. Conclusion: Both models were found to have an appropriate performance. However, considering accuracy as an important criterion for comparing the performance of models in this domain, it can be argued that SVM could have a better efficiency than LR in diagnosing OSA in patients.

Data represents a compendium of information that perpetually expands with each passing moment, contributed by individuals worldwide. Within the domain of medical science, this reservoir of data accumulates at an almost exponential rate, doubling in volume annually. The emergence of advanced Machine learning tools and techniques, subsequent to a substantial evolution in data mining strategies, has bestowed the capacity to glean insights and discern concealed patterns from vast datasets, thus enabling extensive analytical pursuits. This study delves into the application of Machine learning algorithms to enhance societal well-being by harnessing the transformative potential of Machine learning advancements in the domain of blood glucose concentration estimation through Regression analysis. The culmination of this investigation involves establishing a correlation between glucose concentration and hematocrit volume. The dataset employed for this research is sourced from clinically validated electrochemical glucose sensors (commonly referred to as glucose strips). It encompasses diverse levels of both glucose concentration and hematocrit volume, the latter being furnished by an undisclosed source to ensure copyright compliance. This dataset comprises four distinct variables, and the aim of this research involves training the dataset using Regression techniques to predict two of these variables. Our results indicate that when utilizing linear Regression, the R2 score for GC is approximately 0.916, whereas for HV, it reaches around 0.537. In contrast, employing the Support Vector regressor yielded R2 scores of about 0.961 for GC and 0.506 for HV.

In this article, an approach for fitting a fuzzy linear Regression model based on Support Vectors is presentedwhen the response variable, model parameters and errors are considered as fuzzy numbers.In this method, the objective function is based on the sum of the absolute values ​​of the distances of the hypothetical points to the non-parallel border hyperplanes. The presented model has good robustness to the presence of outlier data. The proposed model has been compared with some other models based on three goodness of fit indices.

In Machine learning, models are derived from labeled training data where labels signify classes and features define sample attributes. However, noise from data collection can impair the algorithm’s performance. Blanco, Japón, and Puerto proposed mixed-integer programming (MIP) models within Support Vector Machines (SVM) to handle label noise in training datasets. Nonetheless, it is imperative to underscore that their models demonstrate an observable escalation in the number of variables as sample size increases. The nonparallel Support Vector Machine (NPSVM) is a bi-nary classification method that merges the strengths of both SVM and twin SVM. It accomplishes this by determining two nonparallel hyperplanes by solving two optimization problems. Each hyperplane is strategically po-sitioned to be closer to one of the classes while maximizing its distance from the other class. In this paper, to take advantage of NPSVM’s fea-tures, NPSVM-based relabeling (RENPSVM) MIP models are developed to deal with the label noises in the dataset. The proposed model adjusts observation labels and seeks optimal solutions while minimizing compu-tational costs by selectively focusing on class-relevant observations within an ϵ-intensive tube. Instances exhibiting similarities to the other class are excluded from this ϵ-intensive tube. Experiments on 10 UCI datasets show that the proposed NPSVM-based MIP models outperform their counter-parts in accuracy and learning time on the majority of datasets.

In various researches, implementation of meteorological parameters in drought prediction is studied. In the current work, meteorological drought classes based on Standardized Precipitation Index (SPI) for six seasonal scenarios (autumn, winter, spring, autumn +winter, winter +spring, and autumn +winter +spring) and meteorological predictors contained ground and sea surface temperature, weather temperature (at 300, 500, 700 and 850 mi bar) and geopotential height (at 300, 500, 700 and 850 mi bar) wide of North (0, 60) and East (0, 90) was applied in prediction models based on data from 1975 to 2005. In these models, temporal range of meteorological predictors is between Octobers to April month on the same predicted SPI. SPI was calculated based on mean precipitation at seasonal time scale in the main watershed of Tehran (Taleghan, Mamloo) by verse Weighted Distance method. The well-known statistical supervised Machine learning method, Support Vector Machine (SVM), is applied to predict SPI. Regarding to selected data points, the effective regions on Tehran precipitation are southern, southwestern and northwestern of Iran in spring, northern and northwestern in autumn and northwestern and western in winter. SVM depicted accurate results in prediction of SPI, spatially prediction of SPI in all scenarios, and it can be proposed as a very suitable statistical learning method in investigating of nonlinear behavior of meteorological phenomena with a short samples. The predicted SPI in spring and autumn are more accurate than the other scenarios.