support vector machines

Intrusion detection is a classification problem in which various machine learning (ML) and data mining (DM) techniques are used to classify network data in normal traffic and attack. In addition, the types of network attacks have changed over the years. This paper tries to compare two models of intrusion detection systems, which include adaptive neuro-fuzzy inference systems (ANFIS) and support vector machines (SVM). In addition, it examines and evaluates several instances of data sets related to intrusion detection systems. In the following, a new hybrid method is proposed that uses Particle Swarm Optimization (PSO) to create a classifier combination to provide better accuracy for intrusion detection. Experimental results show that the new method can produce a better performance based on different evaluation criteria. This paper lists the different datasets for evaluating the IDS model and discusses the performance of the proposed hybrid method on the IDS datasets that can be used to efficiently and effectively use the datasets to develop IDS based on ML and DM.

Fog is a cloud that closes to the ground. The components of fog and cloud complement each other. These components provide mutually beneficial interdependent services for communication, processing, control, and storage across the network. Attacking the fog nodes are as important as attacking the cloud. Since the fog node has more limited resources, it is more targeted by intruders. In addition, fog nodes are more attractive to attackers because they have less computing power and are located closer to the attacker than the cloud. But the key point is that access to limited resources makes it easier to save the fog node because the fog does not have the complexities of the cloud, and it is easy to run an intrusion detection system on it. In this article, focusing on the resource limitation in the fog node, we will invent a method to save the fog node. In the proposed method, the support vector machines (SVMs) technique is used. Among the advantages of using the support vector machine, we can mention not being trapped in local optima, solving the over fitting problem, and ease of working with high-dimensional data. Based on the research, support vector machine is the most widely used machine learning method for Internet of Things security articles in the literature. In this article, in order to conduct tests, according to published global statistics, the most important category of web attacks, i.e. SQL injection attacks, is considered. The average detection accuracy is obtained and the results of the evaluations indicate the acceptable efficiency of the proposed method.

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.

Given the inefficiency of static analysis methods due to malware techniques such as code polymorphism, metamorphism, and obfuscation, and self-modifying code, leveraging dynamic and heuristic analysis   methods that are based on the analysis of runtime behavior of malwares, have become particularly important. Environment-aware malware that attempts to conceal its malicious behavior through dynamic anti-analysis methods has caused problems for dynamic analysis detection methods in practice. The purpose of this study is to present an effective method for environment-aware malware detection. Regarding to split–personality of such malware behaviors, this research has proposed an effective way to detect environment-aware malware. This method is based on system call monitoring of malicious and benign  samples under the two NtTrace and drstrace softwares with different monitoring techniques and calculating behavioral distances as training data to create a Support Vector Machine model. Finally, the resulted    support vector machine classifier is used to detect this type of malware with an average precision, recall and accuracy up to 100%, whereas the evaluation of previous related work shows an average precision, recall and accuracy 96.85%, 95.68% and 96.12%, respectively.

In recent decades, increasing population density and economic and industrial activities in metropolitan cities has increased traffic volumes and, consequently, increased levels of air pollution. The major source of air pollution in major developing cities is the massive transport of vehicles that use more than standard fuel and energy, and heavy traffic in the streets of these cities is often rooted in problems such as there is a lack of traffic management and traffic culture. One of the important issues in cities and metropolises that face pollution problems and harmful effects is the issue of informing about the future status of air quality and the amount of urban air pollution to the people. This can be achieved through daily or even hourly forecasts of air pollution and preventing people from being exposed to contaminated areas and their irreversible consequences. Therefore, the need to predict the quality of the air and the quantitative estimates of the concentration of pollutants in the aftermath of the equipment makes it felt that in this study, the problem of the predicted hourly concentration of particulate matter (PM2.5) in the district 11 municipalities of Tehran have exceeded 80% of the contaminated days under the influence of this pollutant. The difficulty and uncertainty associated with estimating and predicting the share of road traffic volume at the general level of air quality is the most important factor that can, if properly diagnosed, be very helpful. In order to take into account the effects of varying the volume of different traffic fleets in the process of changes in the concentration of pollutants and air pollution, it is necessary to pay attention to the effects of other influential variables including hydrological variables, geographical variables, etc. To achieve this, The methods of analytic analysis seem to be able to examine all of these effects together and in an omnipresent manner. The method used to predict this study is one of the methods for analyzing neural networks called Support Vector Machine (SVM). Artificial neural networks are important tools in the field of computational intelligence. Different types of artificial neural networks have been introduced, mainly in applications such as classification, clustering, pattern recognition, modeling and approximation of functions (or regression), control, estimation and optimization of the case Are used. Support Vector Machines (SVM) are a special type of neural network that, unlike other types of neural networks (such as multi-layer perceptron MLP and radial base functions of the RBF), instead of minimizing the error, minimize the operational risk of classification or modeling. Slowly This tool is very powerful and can be used in various fields such as classification, clustering and regression. The results of this study showed that SVM models work well in predicting the contribution and time share of road traffic in propagation of particulate matter, and predictions are well-coordinated with observations. It provides the opportunity to be used as an air quality management tool. Variable significance analysis results for SVM models provide this opportunity to be used as a tool for air quality management, in which the sensitivity of models to variations in emissions can be used to evaluate the effectiveness of a The air quality management scenario will test traffic fleet technology, combine the traffic fleet or its volume.

Knowing the quantity of generated solid waste play a very significant role in solid waste management programs in a region. Due to lack of measured data as well as unavoidable errors in measurements, assessment of volume of generated solid waste is always challenging. Also, field measurement and continues monitoring of the volume of solid waste is usually costly, difficult and time-consuming. Accurate prediction of solid waste generation can be regarded as a key factor in future solid waste management system planning. Conventional forecasting methods in solid waste generation forecasting frequently use the demographic and socioeconomic factors in a per capita basis. In most cases, insufficient funds, the limited measuring equipment, lack of appropriate management systems and due to the lack of recorded data for the volume of generated solid waste cause many problems in integrated solid waste systems management (Dyson and Chang, 2005). In this study, three computational intelligence techniques including M5P model trees, support vector machines (SVM) and multi-layer perceptron (MLP) artificial neural network are utilized to predict solid waste generation in Hormozgan Province, Iran. After a sensitivity analysis, four more influential factors including elevation, population, urban development index (measures the level of development in cities based on infrastructure, the municipality established year, the metropolitan area, population, city product and income, health and education) and the frequency of garbage collection were considered in developing models. The performance of proposed models in solid waste generation forecasting are assessed via different error evaluation indices and finally the results are compared.

Abnormal waves are extremely large and unusual which rarely occur, but cause serious damages. Various factors such as extreme storms, particular topography of the seabed, marine currents, wave-wave interaction with different wavelengths and frequencies may cause wave occurrence and transformation. The main objective of this study is to propose a new hybrid approach to predict the occurrence of abnormal waves using wavelet transform and support vector machines (SVM) classifier based on meteorological data. The data sets used in this paper are from two major hurricanes Dean 2007 and Irene 2011 at four locations namely, 41004 and41041 in the Gulf of Mexico. To predict the occurrence of abnormal waves, at the first extreme waves are detected using wavelet transform. The outputs of this method are considered as SVM classifier targets. Wavelet transform is applied on the significant wave height data samples. The abnormal waves are readily identified from the wavelet spectrum as an area of high energy. The inputs of SVM classifier models are historical metrological data, including: Wind direction (WDIR), Wind speed (WSPD), Sea level pressure (PRES), Air temperature (ATMP), and Sea surface temperature (WTMP). The experiment results show that the proposed method is able to predict the occurrence of extreme wave heights with height accuracy.

In this research, a system is proposed for detecting fertility of eggs. The system is composed of two parts: hardware and software. The fabricated hardware provides a platform to obtain accurate images from inner side of the eggs, without harming their embryos. The software part includes a set of image processing and machine vision processes, which is able to detect the fertility of eggs from captured images, without any sensitivities to different types of eggs (e.g. with different thickness of the eggshell). In order to classify the fertile and infertile eggs, two classifiers based on Artificial Neural Networks (ANN) and Support Vector Machines (SVM) are designed and tested. It means that, to have a fully automatic fertility detection machine, we design two machine learning approaches using SVMs and ANNs to classify fertile and infertile eggs. That is, instead of using a predefined threshold values for distinguishing fertile pixels of egg images from infertile ones, we try to train the machine to do the job automatically. After training the machine using both classification algorithms, the performance of them are accurately investigated and measured in order to select the appropriate one. To evaluate the system, an egg image dataset is provided including 1200 images captured from incubated eggs. Extensive experiments are performed using the provided dataset, which confirm the reliable performance of the system. Comparisons with other fertility detection approaches applying different methods and algorithms confirm that the proposed machine outperforms more complex systems. Performance evaluations of the two proposed classifiers confirm that the SVM based classifier, with average detection accuracy of 50.57% at day 1 of incubation, 83.67% at day 2, 94.20% at day 3, 98.03% at day 4, and 98.91% at day 5, performs better than ANN based classifier, and it is also less sensitive against the reductions in training samples, which can be a serious issue when we are not able to provide more training samples.

Any industry needs an efficient predictive plan in order to optimize the management of resources and improve the economy of the plant by reducing unnecessary costs and increasing the level of safety. Rotating machinery is the most common machinery in industry and the root of the faults in rotatingmachinery is often faulty rolling element bearings. Because of a transitory characteristic vibration of bearing faults, combining Continuous wavelet transforms with envelope analysis is applied for signal proseccing. This paper studies the application of independent component analysis and support vector machines to for automated diagnosis of localized faults in rolling element bearings. The independent component analysis is used for feature extraction and data reduction from original features. The principal components analysis is also applied in feature extraction process for comparison with independent component analysis does. In this paper, support vector machines-based multi-class classification is applied to do faults classification process and utilized a cross-validation technique in order to choose the optimal values of kernel parameters.

Rainfall-Runoff is considered one of the most important processes in water resources studies. In this study, to simulate the daily rainfall-runoff process of Balikhluchay Basin, four hybrid models of Support Vector Machine, Artificial Neural Networks, Wavelet-Support Vector Machine, and Wavelet-Neural Networks have been applied and compared. Daily Rainfall-runoff data for the period of 2000 to 2008, have been used for training and testing the models. In general, the results indicated acceptable accuracy of all the models. In terms of priority, the hybrid model of Wavelet-Neural Network with the highest accuracy and lowest errors was in the first rank and the Hybrid models of Wavelet-Support Vector Machines, Artificial Neural Networks and Support Vector Machines, were in next priorities.