جستجوی مقالات مرتبط با کلیدواژه "شبکه عصبی" در نشریات گروه "علوم اجتماعی"

Recognizing one's identity is a vital thing for humans, since long ago, humans needed to distinguish friend from enemy for immortality. Therefore, today, attempts have been made to mechanize identification systems.The face is one of the most important biometric components of a human being, through which useful information such as race, identity, age, gender, and facial expressions can be extracted. Automatic face recognition and analysis is one of the most important and challenging tasks in the object recognition branch, which has many applications in human-computer interaction, human-society, psychology, and security issues.in all the identity recognition articles, the data is a combination of male and female genders, and in these articles, only face recognition or gender recognition methods are used for identity recognition.In this report, in order to increase the speed and increase the accuracy rate of its identity recognition system, we combined two gender recognition methods using n special face vectors and SVM and face recognition using m facial features and neural network. In this way, first, using SVM, he separated the data into two groups of male gender and female gender, which is very effective in reducing the activity of the next step, that is, recognizing the identity of people by recognizing the faces of one gender, by the neural network.

Undoubtedly, human societies have always faced the phenomenon of financial crimes as uninvited guests and are always exposed to harm. This research has been done using data mining tools and with the aim of increasing the efficiency and accuracy of financial crimes, with neural network and decision tree algorithms. Statement of Problem: The forthcoming issue indicates that the spread of cybercrime and, in parallel, the increase of recorded crime data, has confronted us with a large amount of data, each of which contains a large number of messages and information. If analyzed correctly, this data can help police organizations identify and track, as well as predict and prevent crime.

This study improves the efficiency and accuracy of financial crimes by data mining more than 4500 records of cyber financial crimes using data mining tools. The type of data mining method in this article is predictive. This study, by analyzing the data and crime variables taken from the FATA police database, using decision tree algorithms and neural networks, has created a new method to increase efficiency and accuracy in financial crimes, so that efficiency using From the neural network algorithm in the proposed method is 70. 11% and in the decision tree algorithm is 87.78%.

Using the existing tools in the field of data mining, it is possible to detect crimes or prevent their occurrence by using appropriate and optimal human resources of the police. It is hoped that police organizations can be assisted in identifying criminals and, more importantly, in predicting and preventing cybercrime, especially financial crime.

Motorcyclists due to lack of protection at the scene of the crash and also being less exposed in the eyesight of other vehicles are more vulnerable for sever crashes. The main purpose of this study, is to identify the factors affecting the severity of motorcycle crashes.

Neural network, random forest and statistical analysis (comparison of the average severity of accidents in variables) have been used in this paper. To conduct this research, two series of data (accident data and behavioral data collected from motorcyclists in Tehran based on field interview) have been used.

The severity of crashes increases with the rise of the frequency of speeding, passing red lights, using mobile phones, performing dangerous movements, not usign helmet and crossing the pedestrain pathway. However, with the increase in the health status of motorcyclists and the technical appearance of the motorcycle the severity of accidents will decrease. Among the districts of Tehran, district 6 had the highest frequency of crahses. Neural network model, is able to predict fatalities, injuries and propoerty damage crashes with 24, 76 and 51 percent accuracy. Furthermore, in the random forest model, the accuracy of predicting fatal, injury and property damage crashes are 53, 69 and 32 percent.

It indicate that the traffic violations of the motorcyclists are the main reasons for crash severities. Thus, it is necessary to have an overlook about performing more restrict traffic fines for this group. In addition, becasuse of the importance of the individual health condition and technical appearance of the motorcycle it is necessary to determine rules for controling them with issuing cycling licence and other intervals after that. Finally, the high rate of motorcycle crashes in the district of 6 in Tehran can be related to high capacity rate of transit  in the zone.

Identifying and analyzing crucial parameters that cause accidents in highways,can help improving traffic. This paper addresses a multi-parameteroptimization problem in order to identify the parameters affecting the severityof accidents on highways in Tehran city and uses a combination model ofNeural Network and Genetic Algorithm to perform the analysis. The methodof this research is descriptive-cross-sectional. The statistical population of thisstudy is the accidents data on the highways of Tehran during 2015-2016. Inthis research, it is attempted to identify and prioritize the parameters affectingthe severity of accidents in Tehran city using a combination model of neuralnetwork and genetic algorithm. For this purpose, in the hybrid model, theseverity of the accident is considered as the dependent variable and the fourgeneral categories of variables namely climate, road, vehicle and driver areconsidered as independent variables. Then, using artificial intelligence methodand data preprocessing, the optimal structure of the neural network model wasdetermined and finally, the result of the neural network model was consideredas the input of the genetic algorithm. The results not only determines andprioritizes the main parameters affecting the severity of accidents inTehran(including: 1- driver behavior, 2- how the vehicle is moving, 3- type ofvehicles and 4- highways safety status), but also Indicates that the combinationmodel of Neural Network and Genetic Algorithm has a good performance inidentifying the parameters affecting crash severity in Tehran, And couldprovide a new insight into designing a pattern to understand better and preventfuture accident-related accident injuries.

One of the effective factors in the reduction of the accidents is the improvement of the friction of the existing pavement surface structures, which is often measured with a conventional parameter against skid. Resistance to skid is an important feature of road pavement surface, which has a significant role in providing safety and prevention of accidents resulting from skidding. The aim of this study is to investigate the effect of various contaminants on the skid resistance of asphalt mixture.

There are various pollutants that affect the skid resistance of pavements, such as dust, dune sand, oil and diesel, exhaust fumes, etc. Present research has taken in account the interpretive and comprehensible results from the neural network, which is a data mining method used in predictions and statistical modeling.

The results of this research can be used to select the type of asphalt mix according to the weather conditions of each region. As it is evident, in the desert areas, there is a lot of dune sand on the road surface, which reduces the resistance of the skids or in busy transportation areas, smoke due to the exhaust of trucks, as well as the dust in the air settle on the road surface and reduce the skid resistance.

The data obtained was analyzed using artificial neural network method and it was observed that the sort of contaminants had the most impact on the variation of the skid resistance of the samples. Field observation and installation of electrical equipment in high contamination sites and inspections were performed, and for the reduction of accidents it was recommended that accident prone areas should be eliminated before the first rainy season after the dry seasons, in order to avoid the dangers of skidding.

Accurate traffic forecasts has important role in providing real-time data traffic, reduce congestion on the roads and improve traffic safety. In this study, a combination of multi-layered back propagation neural networks to predict the peak hour traffic flow with wavelet transform is used. According to the proposed method, using the wavelet transform, a pre-processing is done on the data volume traffic flows to obtain more detailed information about the dynamics of the problem; then as intelligence, training and testing data is presented to the neural network. Trained network using validated assessment functions and to predict the peak hour traffic volume may be used during the next week. Predictions made to measure, traditional back propagation neural network design and compared the results with the proposed method. The results show that the proposed method, the peak hour traffic flow with higher precision than conventional neural network predicts. The method proposed in this study has been done on the basis of parameters and can be used in practical applications.

Given the importance of role of women in today’s world, improvement of their employment and guiding them through labor market will provide an opportunity for society to benefit from this great potential work force. Thus identification of factors that make the conditions of cooperation of women in economic and social areas has main priority. Present study investigates the impact of some variables including Percentage of girls enrolled in primary, secondary and higher education, share of value-added of industry in GDP (development process indicator), square of the value added of industry (development index), marriage to divorce ratio (socio – cultural index), government share of GDP (size of government index) on women employment in 1968 to 2011. This research uses artificial neural network to analyze data. Findings indicate that in this time period increase in development of the country and proliferation of women education are the most influential factors among women as work force.

Road traffic volume which is the number of vehicle travelling from specified place of road at specified time which is the determining factor of available network that using demand and future infra structure of countrie,s road and their available facilities. The main objective of this article is identifying and predicting road travelling by using artificial nervous network. Used network is for problem modeling by multi layer perspetron nervous network. Another application of it is reconstructing not collected information of travelling counting sets with using pervious data and new data.

Negligence to the traffic rules and law breaking regarded as one of the causes of traffic accidents and problems always in all societies that depend on cultural, social, economical and geographical context and its extent and kind will different. Although the role of enviromental phenomena is important but the role of human is the effective factor in decreasing and removing other errors and shortages. This article through practical data searching on the traffic violations data to has studied the ways of recognizing high risk drivers and if it is possible, through knowing the violations records and environmental and personal features, to identify them and ratify traffic rules for reducing traffic violations. In the way of applying categorization method in data searching two techniques including nervous networks and decision tree in violations information have been used. And in the end, to improve the results, combination approach including categorization and decision tree techniques have been used.