machine learning algorithms

The aim of this paper is to design and implement a smart system for detecting electrical leakage current using machine learning algorithms. Leakage current in electrical systems is a major issue concerning safety and efficiency, and timely detection can prevent accidents and irreversible damage. In this study, a dataset of current and voltage from various electrical systems was collected, and then data analysis was performed using machine learning algorithms such as decision trees, neural networks, and support vector machines (SVM). The results showed that the proposed system has the ability to accurately detect leakage currents and can be used as a practical tool in electrical protection and control systems. This approach, in addition to improving detection accuracy, also reduces maintenance and repair costs for electrical systems.

Traditional static geomechanical models are widely used for analyzing stresses in oil and gas reservoirs. However, when these models are employed to investigate heterogeneous stress conditions, challenges such as stress field rotation and errors arise, particularly observed in salt intrusion masses, salt domes, and reactivated faults due to gas injection. Therefore, predicting stress variations in static geomechanical models becomes impractical, especially in deep and ultra-deep reservoirs, heavy oil reservoirs, gas injection reservoirs, and enhanced oil recovery operations. Hence, there is a need for a dynamic physical simulation model to examine four-dimensional stresses, strains, and reservoir deformations. Therefore, the objective of this paper is to develop a coupled geomechanical simulation model to simulate variations in stress, pore pressure, strain resulting from hydrocarbon depletion, subsidence, and reservoir deformation. To achieve this goal, static and dynamic data from oil fields located in one of the oil-rich regions have been utilized. Initially, data preprocessing and determination of geomechanical units were performed using machine learning algorithms. Subsequently, a one-dimensional geomechanical model including rock strength parameters (unconfined compressive strength, tensile strength, friction angle), deformation parameters (Young's modulus, Poisson's ratio, shear modulus, bulk modulus), and pore pressure were constructed, and residual stresses were determined using elastic poroelastic relations. Furthermore, a three-dimensional static, dynamic, and geomechanical simulation model was developed. This model utilized a finite difference reservoir simulator for evaluating the thermo-hydro-chemical (THC) process using ECLIPSE software, alongside a finite element reservoir simulator for thermo-hydro-mechanical (THM) process calculations using VISAGE software, unilaterally coupled in the PETREL 2018 software environment. Consequently, several results including stress, strain, pore pressure changes, and reservoir deformation over ten years from 2021 to 2031 for the Bangestan reservoir were obtained

Corona virus, Severe Acute Respiratory virus and swine flu is a disease caused by acute respiratory syndrome. These viruses require advanced tools to identify dangerous mortality factors with high accuracy due to their immediate spread among humans. Machine learning methods directly address this issue and are essential tools for understanding and guiding public health interventions. In this article, machine learning is used to investigate demographic and clinical significance. The investigated characteristics include age, gender, fever, countries and clinical details such as cough, shortness of breath, etc. Several machine learning algorithms have been implemented and applied on the collected data, the K-Nearest Neighbor algorithm works with the highest accuracy (more than 97%) to predict and select features that correctly represent the status of viruses.

Nowadays, due to increasing importance of energy consumption, many methods have been applied aiming accurate. One of these methods is to use artificial intelligence. Many algorithms are used to do this; Which provide a more accurate result than other building energy simulation tools. This article, reviews some research that has been done in recent years, and related to this issue; To identify the most accurate and most widely used algorithms in this field; The algorithms have the lowest error rate, the highest speed and the most accurate calculations. When using different algorithms, the strengths and weaknesses of each are identified. Researchers have used various algorithms to predict the energy consumption of buildings. The purpose of this article is to better understand the features of different algorithms according to the type of their use; and introduce the best and most widely used machine learning algorithms in recent years. the most widely used methods in order to optimize energy consumption and predict the energy performance of buildings have been ANN, SVM, RF, GA and LR respectively. In this research, a layered framework has been used to select articles; In the first layer, articles that have been published in the recent years were considered; In the second layer, these articles had the keywords we wanted and were in line with our achievements; In the third layer, the subject of the articles was in line with our subject and machine learning algorithms were used to predict or optimize the energy consumption of building.

Definition of the efficient ore-forming processes which are considered as mineralization controls is a fundamental stage in mineral prospectivity mapping. In this contribution, five targeting criteria of geochemical, geophysical, geological, structural and hydrothemal alteration data related to epithermal and Carlin-type Au deposits in Takhte-soleyman district, NW Iran, were integrated. For creation of multi-element geochemical layer, principal component analysis was firstly conducted on stream sediment data of 8 selected elements and it was found that PC2 is the representative of Au-As-Sb elemental association in the study area. Then, 10 fuzzified efficient evidence layers were selected based on area under the curve (AUC) of success-rate curves and prepared for generation of predictive models of Au mineralization. For this purpose, two supervised machine learning algorithms, namely multi-layer perceptron (MLP) neural network and support vector machine (SVM) with RBF kernel were used. Comparison of the generated models demonstrates that the latter is more succeeded in delineating exploration targets than the former one.

Domain generation algorithms (DGAs) are used in Botnets as rendezvous points to their command and control (C&C) servers, and can continuously provide a large number of domains which can evade detection by traditional methods such as Blacklist. Internet security vendors often use blacklists to detect Botnets and malwares, but the DGA can    continuously update the domain to evade blacklist detection. In this paper, first, using features engineering; the three types of structural, statistical and linguistic features are extracted for the detection of DGAs, and then a new dataset is produced by using a dataset with normal DGAs and two datasets with malicious DGAs. Using supervised machine learning algorithms, the classification of DGAs has been performed and the results have been compared to determine a DGA detection model with a higher accuracy and a lower error rate. The results obtained in this paper show that the random forest algorithm offers accuracy rate, detection rate and receiver operating characteristic (ROC) equal to 89.32%, 91.67% and 0.889, respectively. Also, compared to the results of the other investigated algorithms, the random forest algorithm presents a lower false positive rate (FPR) equal to 0.373.

Nowadays, artificial intelligence methods have been broadly developed and applied for variable estimation to facilitate decision making in many fields. Grade estimation is an important issue in evaluating mineral deposits. Geostatistical methods are among the most commonly used approaches for variable estimation. Since these methods are somewhat defective in relation to limited numbers of dispersed nonlinear data, in this study, the support vector regression, a machine learning method, has been used for grade estimation in Esfordi phosphate deposit. The modeling accuracy was 84% according to the test data. Based on the results obtained from the modeling using the support vector regression method, grade estimation has been made within the block model in Esfordi phosphate deposit. The proposed potential areas in the block model can be taken as the the additional borehole sites in the further exploration stage. The tonnage-grade model was also prepared based on the results obtained by using the support vector regression modeling procedure. For example, based on this model, for a 6% cutoff grade, the reserve is about 15.36 million tons with an average grade of 13.59%.

Heart is one of the most important members of the body, and heart disease is the major cause of death in the world and Iran. This is why the early/on time diagnosis is one of the significant basics for preventing and reducing deaths of this disease. So far, many studies have been done on heart disease with the aim of prediction, diagnosis, and treatment. However, most of them have been mostly focused on the prediction of heart disease. The purpose of this study is to develop models for heart disease diagnosis using machine learning, neural network, and deep learning algorithms. The models have been developed using the Cleveland heart disease dataset from University of California Irvine (UCI) repository. After complete data processing, including outlier detection, normalization, discretization, feature selection and feature extraction, the dataset is transformed into two normalized data and discretized data, according to the nature of the algorithms. Moreover, in constructing models of machine learning and neural networks, two randomized searches with cross-validation and grid search with Talos scan approaches are used for model tuning. Among evaluated models, including decision tree algorithms, random forest, support vector machine (SVM) and XGBoost, the highest accuracy is 92.9% using SVM, and among neural network models, multilayer perceptron (MLP) has resulted in the highest accuracy of 94.6%.