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. 

The rock mass permeability is one of the most important parameters regulating to the groundwater flow through the fracture’s rocks. The permeability distribution is an important part of estimating inflow into tunnels. The common methods to rock mass permeability estimation such as lugeon tests are expensive and very time consuming. The use of intelligent methods to estimate or classify data, especially in engineering problems, has been common in recent decades. Many algorithms have been designed and optimized for this purpose. Support Vector Machines (SVM) is one of these methods. In this paper, using the SVM method, the Amirkabir tunnel has been classified from the permeability point of view. In order to optimize the parameters of this algorithm, random search method has been selected. The results show that the accuracy of modelling using this method based on experimental data is around 94.59%. Based on this result, amount 85% of tunnel length is classified in the low permeability category and water inflow into tunnel from this part of tunnel is negligible

Aims: Information of the protein structure is essential to understand the protein functions. Flexibility is one of the most important characteristics related to protein functions. Knowledge about flexibility of the protein structures can be helpful to improve protein structure prediction and comprehend their function. This study was conducted with the aim of investigating the flexibility prediction of protein structures, using Support Vector Machine. Materials & Methods In this study, a balanced dataset containing 95 proteins was used. The features used in the present study for modeling amino acids formed a 33-dimensional Vector. Some of them were obtained by crawling a window with the length of 17 focusing on the target amino acid on the protein chain, and some were only related to the target amino acid. To define the flexibility factor, the characteristics based on the information derived from the twodimensional angular variations was used. The information was calculated for each amino acid by considering the position of each amino acid alone and for the adjacent amino acid pairs in a seventeenth window, and the Support Vector Machine method was used for prediction. Findings The accuracy was 73. 1%, F-measure was 71%, precision was 73%, and sensitivity was 73. 2%. Acceptable superiority of the proposed method was confirmed in comparison with the current methods. The angular representation of each protein was able to accurately demonstrate the 3D characteristics and properties of the protein structure. Conclusion The accuracy is 73. 1%, F-measure is 71%, precision is 73%, and sensitivity is 73. 2% and angular aspect is the best descriptor for flexibility prediction. Angular representation of each protein can accurately reflect the 3D characteristics and properties of the protein structure.

Human users can easily divide a bibliographic reference to its constructing fields such as authors, title, journal, year, etc. However, due to the variations in formats and errors made by the authors in citing documents, it is difficult to automate this task. There exist many solutions for this problem, known as citation parsing problem in the literature, however, none of them is compatible with Persian language. This is mainly due to high language-sensitivity of these solutions. Considering the important role of citation parsing in tasks such as autonomous citation indexing and information retrieval, in this paper, we propose an intelligent method for citation parsing in Persian language. The proposed method uses the Support Vector Machine (SVM) classification method as its core. The results of testing the proposed method using a dataset designed for this task show 95% in average for precision, recall and F1 measures for extracting different fields from a bibliographic reference which is quite plausible.

Purpose: The aim of this paper is to present an enhanced variant of Twin Parametric-Margin Support Vector Machine (TPMSVM) that improves classification performance. Methodology: By replacing a variable in the objective function, we keep the samples of one class farther from the parametric margin hyperplane of the other class. Findings: The enhanced model is convex for both linear and nonlinear cases. Also, numerical experiments on UCI datasets show that the enhanced model performs better compared to two similar models for both linear and nonlinear cases. Originality/Value: The previous studies of TPMSVM that increased the accuracy through approaches such as assigning weights to data sample, converting it into an unconstrained model and adding a new term in the objective function, did not guarantee that all samples will be far and on the negative side of the margin hyperplane. However, this study provides an approach to overcome this disadvantage of TPMSVM.