Search

The ionosphere is the upper part of the Earth's atmosphere, which is considered to be approximately 70 to 1000 km above the Earth's surface. Ionosphere modeling has been one of the goals of spatial geodesy since 1970. In many ionosphere modeling using satellite measurements such as GPS, total electron content (TEC) are used as observational input data. In recent years, modeling and prediction of the TEC have been considered by researchers with methods that have high speed and accuracy. One of the branches that has been able to show good capabilities in the field of estimation and modeling is machine learning methods (ML). Machine learning includes fuzzy inference systems (FIS), artificial neural networks (ANNs), genetic algorithms (GAs), support vector machines (SVMs), and evolutionary communications (ECs). Since 1993, with the advancement of computer technology, many new and hybrid algorithms, such as the adaptive neuro-fuzzy inference system (ANFIS), have been developed in ML. Another new effective approach in ML is the support vector regression (SVR) method. The SVR is a kernel-based ML method for classification and regression in which the risk of incorrect classification is minimized. The structure of an SVR network has a lot in common with the ANN, and the main difference is practically in the way of the training algorithm. In general, this method is divided into linear and nonlinear modes. In this paper, the TEC of the ionosphere is modeled and evaluated with ML models. Support vector regression (SVR) and artificial neural network (ANN) methods are used for local TEC modeling. In both models, the latitude and longitude of the GPS stations, day of the year (DOY), hours, AP, KP, DST, and F10.7 are considered an input vectors. Also, the value of VTEC is considered as the output of the models. The main innovation of this paper is in evaluating the effect of different physical parameters on the accuracy of ML models. Using observations of 15 GPS stations in the northwest of Iran from 193 to 228 in 2012, new models are evaluated. Also, the results of the new models are compared with the results of the global ionosphere map (GIM), the IRI2016, and NeQuick empirical models in two internal and one external control station. Statistical indices of root mean square error (RMSE), relative error, dVTEC, and correlation coefficient are used to evaluate the error of the models. Sensitivity analysis of SVR and ANN models to input parameters is performed and the importance of each physical parameter in spatio-temporal modeling of the ionosphere is investigated. The results obtained from this paper show that in both high and low geomagnetic and solar activities, the SVR model in internal control stations has a higher accuracy than other models. But at the external control station, the error of the SVR model is much higher than other models. Determining the parameters of the kernel function using observations at the territory of the studied network is the reason. Also, the sensitivity of SVR and ANN models is increased to the physical parameters F10.7, KP, DST, and AP, respectively. For precise local ionosphere modeling, the effect of these parameters must also be considered. [ABSTRACT FROM AUTHOR]

In a wireless sensor network (WSN), due to various factors such as limited power, sensor transferability, hardware failure and network problems such as packet collisions, unreliable connection and unexpected damage, the amount sensed to the header or base station is not Arrives. Therefore, data loss is very common in wireless sensor networks. Loss of measured data greatly reduces WBAN accuracy. Because WBAN deals with the vital signs of the human body, network reliability is very important. To solve this problem, missing data must be estimated. In order to predict the missing values, a model for estimating lost data based on LSTM (short-term memory) neural network is presented in this paper. This model combines five vital signs as input to predict the amount lost. The results show that sgdm-LSTM is a good way to estimate the amount lost. In addition, experimental results show that the mean square root error of the estimated value is lower than other methods. This value is4.1495 with the best network parameters. [ABSTRACT FROM AUTHOR]

The proper forecasting of the intermittent wind power for optimal power systems operation is a really tough and complicated issue. In this paper, a successful method is proposed for predicting the wind power productions which is based on the self-organized neural networks called Group Method of Data Handling (GMDH). By analyzing and discovering the hidden relationships between the inputs, the GMDH-based neural network intelligently presents the optimal model and predicts the output variable. Patterns used in this study are based on the two methods of artificial intelligence and information theory. At first, the effective variables are selected based on Mutual Information (MI) technique and the mixed particle swarm and genetic algorithm and after that the proposed forecast engine is used. In contrast to the mutual correlation method, in the proposed cross-entropy-based approach of this paper, non-linear relations between the variables are considered and the selection of effective variables in the forecasting of wind power in which nonlinear fluctuations and trends are observed are chosen more precisely and more accurately. In order to evaluate the ability, speed and accuracy of the proposed framework, real-world data of Sotavento wind farm in the Spain were used. The results of the study indicate that the proposed technique has a higher speed and accuracy than other methods. [ABSTRACT FROM AUTHOR]

In this paper, an improved, real-time, highly accurate control-oriented style, named Neuro Mean Value Modeling, is presented for IC engine modeling. This model is a combination of neural networks and mean value model, and is able to overcome the shortcomings of both styles. In other words, taking advantage of both methods, this grey-box extension will be of more reliability than a mere black-box neural network, and also of more accuracy than roughly white-box mathematical relations of mean value model. In this paper, the model is modified to become suitable for designing an engine controller. Thanks to the sophisticated methods applied (such as committee method, improved partitioning, and especially, simplifying neural networks’ tasks), neural networks of high accuracy with line-like regressions will be achieved. As will be seen, the model is precisely validated based on experimental-software data and it is capable of accurately predicting the engine’s outputs (such as pollutant emissions, manifold pressure, knock probability, and engine speed) all in real time. In the end, the effect of engine control inputs on pollutant emissions and fuel consumption will be examined. The engine employed to establish the model is a ported fuel injection SI engine. [ABSTRACT FROM AUTHOR]

In this paper, experimental defect diagnosis and the classification of its size in the outer race of angular contact ball bearing with acoustic emission method and artificial neural network are presented. In an experimental system, bearings are loaded at four speeds of 600, 900, 1200, and 1500rpm with four loads from low to high. Loads are applied to the outer race with the help of four bolts with equal and specific torques. Since the bearing is angular type, this type of loading is converted to radial and axial combined loading simultaneously and differs from conventional loads in deep groove bearings. Acoustic emission waves are recorded using broadband sensors in two states, healthy and defective. Therefore, to diagnose the defect, different states can be compared with the healthy. The spark method was used to create an artificially seeded defect. In analyzing the results, a new parameter called “the total time above threshold” was introduced to increase the efficiency of defect diagnosis and classification of its size. With the help of the introduced parameter and 4 conventional acoustic emission parameters and using an artificial neural network, the performance of the artificial intelligence system was 95.1% in defect diagnosis and 94.4% in defect size classification. [ABSTRACT FROM AUTHOR]

Purpose: This study aims to model the impact of Covid-19’s stress and resilience on job burnout in knowledge-based companies. Design/methodology/approach: This study is typically quantitative and cross-sectional and in terms of purpose it is applied research. The statistical population included the managers of the first and second-tier and the employees of the knowledge-based companies. Based on the equation of the statistical sample size of the unlimited population, 384 were examined. The standard questionnaires of Maslach and Brief Resilient Coping Scale (BRCS) and Covid-19 researcher-made questionnaires were used for data collection. Radial Basis Functions - Artificial Neural Network (RBF-ANN) was used for data analysis. Findings: 65% of the managers and employees of knowledge-based companies were at moderate and lower resilience levels and 61% of the statistical sample had job burnout. Also, the amount of stress caused by Covid-19 was higher among married women compared to others. The RBF method was used to design the ANN. Accordingly, the number of neurons in the input layer was equal to 10, the number of neurons in the single hidden layer was equal to 35, the number of neurons in the output layer was equal to 1, and σ was equal to 10. 70% and 30% of the data were used for training and testing, respectively. In the designed ANN, all but one of the test data, and all but two of the experimental data were correctly predicted and the Root Mean Square Error (RMSE) error was less than 0.3. Finally, based on the obtained results, the proposed model was confirmed. Research limitations/implications: The difficulty of accessing statistical samples in Covid-19 conditions and the resulting limitations along with the lack of relevant research background were among the limitations of the present study. For future research, similar comparative studies are suggested to be conducted in the manufacturing knowledge-based companies for modeling and adapting the results and conducting a study using other methods of ANN design, including multilayer perceptron (MLP). Also, separating the areas of activity of knowledge-based companies and comparing the results are suggested as the subjects of study on the variables of this research. Practical implications: Since in the research related to social sciences and humanities, less use is made of engineering methods such as neural network design, the present study seems innovative in terms of subject and methodology and the researchers and experts who are interested in the subject of this study can benefit from the findings. Business and entrepreneurship and organizational behavior, engineering sciences and sustainability issues, students and managers, and employees of technology and knowledge-based companies are the other beneficiaries of this study. Social implications: Since there is no immediate and definitive solution to reduce the stress and burnout of managers and employees of the startups, constant pressure has created a long-term detrimental situation for startup companies. Addressing this issue is necessary because the performance and productivity of a company require the physical and mental health of its managers and employees; stress and resilience are also the two factors affecting job burnout which have been exacerbated by the Covid-19 crisis over the past two years. Originality/value: Because dealing with complex relationships between research variables requires the use of precise and in-depth analytical methods, in this study, an ANN was used to predict their behavior and the impact of variables on each other. Therefore, the attempt made to reduce the theoretical gap and the contribution made in theory based on innovation in the subject and research variables and the analysis method has led this paper to have an interdisciplinary approach. [ABSTRACT FROM AUTHOR]

In this paper, in order to fault locate in the transmission network, a discrete wavelet transform is used to extract the fault characteristics from the zero sequence current, in order to train the artificial neural network. Initially, Fortescue transform, the zerosequence current seen from both terminals is calculated. By the wavelet transform of the high-frequency information stored in the horizontal component of zero-sequence current from both terminals, and finally by calculating the stored energy in the horizontal components, as well as extracting the maximum scale of horizontal component, we can identify certain features of fault that are suitable for training the neural network. The simulation results show that the horizontal components maximum scale as well as the energy stored in these components strongly depend on the fault resistance, type of fault and fault location. Therefore, educational data should be selected to make these changes well so that the neural network does not suffer from its diagnosis. Finally, the proposed method is implemented on the test grid whose results show the performance of the method with overall accuracy of 98.6% and maximum estimation error of 0.1666%. [ABSTRACT FROM AUTHOR]

Background and Aim: Skin cancer has grown dramatically over the past decades, and the importance of early treatment is increasing day by day. The purpose of this study is to use deep neural networks to create an auto-diagnosis system for melanoma, in which data is directly controlled as part of a deep learning process. Materials and Methods: In this paper, studies on related pictures of skin cancer were performed. For the diagnosis of benign or malignant skin cancer, the deep neural network classifier is used with the help of the Tensorflow framework and the use of the Keras libraries. The dataset which are used in this study consist 70 images of melanoma and 100 images of benign moles. In the proposed model, 80% of the database images are used for training and 20% of the database images are selected for testing. Results: The proposed method offers a higher detection accuracy than other existing methods, which has increased the accuracy of diagnosis in most cases by more than 10%. The high accuracy of the diagnosis and classification and the speed of convergence to the final result are the characteristics of this Research Compared to other Research. Conclusion: An automatic system based on deep learning is presented to identify and categorize skin cancer which provides high accuracy and speed. [ABSTRACT FROM AUTHOR]

In this paper, the effect of surface charges and pollution on flashover voltage of polymeric insulators under DC voltage have been investigated and analyzed. For this purpose, four types of polymeric insulators with different profiles have been selected and then experimental tests of DC flashover voltage have been performed. According to the obtained results, with increasing pollution in presence of positive surface charges flashover voltage reduce, but negative charges increase it. Also in this paper, a model based on geometric parameters including specific leakage distance, shed spacing to shed depth ratio, amount of surface charges and pollution is proposed using an artificial neural network. The comparison of the obtained results from experimental tests and ANN model show that the results of ANN are consistent with experimental tests. [ABSTRACT FROM AUTHOR]

A prerequisite in the majority of control processes is modeling. The model used to design a controller must be both accurate and real-time. Utilizing prevalent approaches of modeling, namely modeling based on (numerically) solving the equations governing the fluid in the combustion chamber, is too time-consuming and not suitable for control purpose. This paper aims to model combustion in an SI engine by means of neural networks and present an accurate and fast-response model for combustion. Obviously, any training procedure of neural networks involves empirical data acquisition. On the other hand, engine testing is highly expensive, and testing data tables available (in industry) is not sufficient to train neural networks. In this paper, first with the aid of CFD software, a one-dimensional model of an engine is constructed, and then calibrated using the factual experimental data at hand. Afterwards, acquiring data required is performed via the validated CFD model. As a matter of fact, because of the lack of access to necessary experimental coefficients, calibration is an extremely complicated and time-consuming process. An attempt is made to accomplish and spell out the calibration of the engine model in the GT-Power software, in a scientific practice. After a brief survey on the methods employed in designing the neural networks, modeling of the combustion chamber will be stated. Eventually, the response of the constructed NN model will be compared to the results gained from the GT-Power software, and the great accuracy of the NN model will be shown. [ABSTRACT FROM AUTHOR]

Introduction: Most heart diseases show symptoms on ECG, but diagnosing heart disease with ECG requires the knowledge and experience of medical specialized. Because these specialists may not always be available, it is necessary to design tools to diagnose heart disease in these situations. Objective: In this paper, a two-stage approach based on artificial neural networks is designed to diagnose heart disease using ECG information. Method: To design the proposed approach, first ECG information for 861 refers to a number of medical centers in Arak city is collected and and data consulted is proccesed by specialists. Then 154 features from ECG as input variables in proposed approach has been specified. In the first stage of approach, an artificial neural network is designed to detect the status of the ECG in two situation as usable and unusable. Then, in the second stage, using the usable ECG information, an artificial neural network is designed to diagnose the presence or absence of heart disease. Finally, the performance of the two-stage approach is evaluated and its accuracy and precision in determining the status of the ECG as well as the disease status is determined. Results In the proposed approach, the neural network for the determining of ECG status has an precision of 97.1% and an accuracy of 97.3%, and also the neural network for the diagnosis of heart disease has an precision of 95.8% and an accuracy of 95.4%. Conclusion: Considering the high effeciency of the proposed approach in the determining of ECG status and also diagnosing heart disease, it is possible to use this approach as a reliable assistant to assist the treatment staff. [ABSTRACT FROM AUTHOR]

Background and Objective: Water distribution networks are prone to terrorist attacks by injecting toxic substances, due to their vastness and availability. The main objective of this paper was detecting the extent of intentional pollution in the urban water distribution network by self-organizing map (SOM). Materials and Methods: The existing hydraulic condition of the water distribution network covered by reservoir No. 4 in Tehran was modeled as a pilot. Possible injection scenarios of contamination in different parts of the water distribution network were performed using qualitative analysis of the water distribution network, using the EPANET analyzer engine and coding in R software environment. Artificial neural network of SOM was used to find the contamination range for the injection of arsenic at different times and places in the distribution network. Results: The concentration of contamination at a certain point decreased over time and a high correlation was observed between time and concentration. The extent of contamination depended on the consumption of subscribers and consequently, the time of contaminat injection. The results of the artificial neural network model showed that the method developed in this research was 91% accurate and was able to determine the extent of contamination in the water distribution network at high speed. Conclusion: SOM can be used as a complement to the water quality monitoring and pollution detection system in the urban water distribution network to determine the extent of pollution when detecting potential pollution in the shortest possible time, and as an alternative to quantitative-qualitative modeling of the water network. [ABSTRACT FROM AUTHOR]

Nowadays in some manufacturing processes, the quality of a product or process is well expressed by both correlated attribute and variable quality characteristics. To best of our knowledge, there is no method for monitoring the covariance matrix of multivariate-attribute quality characteristics. In this paper, we propose a multi-layer perception artificial neural network to monitor multivariate-attribute processes as well as to diagnose the quality characteristic(s) responsible for out-of-control signals. The performance of the proposed method is evaluated through a numerical example from both detection and diagnosis perspectives. In addition, the performance of the proposed neural network in detecting shifts in the variance of quality characteristics is compared with two statistical methods first proposed for monitoring the variability of multivariate quality characteristics and developed in this paper for our problem. The results of numerical example show that the proposed artificial neural network outperforms the extended statistical methods in detecting different out-of-control shifts. The results also confirm that the performance of the proposed neural network in identifying the quality characteristic(s) responsible for out-of-control signal is satisfactory. [ABSTRACT FROM AUTHOR]

Desertification is a phenomenon that has more destructive effects in arid, semi-arid, and semi-humid regions than in other regions. This paper tries to provide a map of the future of desertification in Tehran Province, for futurism in the face of land degradation and desertification. The IMDPA model was used to evaluate land degradation and desertification. To use this model and evaluate desertification, three criteria of groundwater including groundwater depletion, electrical conductivity, and sodium adsorption ratio indices, climate criterion including precipitation, aridity, and drought indices, and land use criteria were selected as key criteria effected on desertification according to regional conditions. Land use index map with IGBP standard and zoning map of other indicators were prepared by IDW method for 2011 and 2016. The maps of land use index and other indices were predicted using the CA-Markov model in TerrSet software, and using the RBF method in artificial neural network toolbox, respectively. Scoring based on the IMDPA model, the maps of indices and criteria maps were prepared for 2011, 2016, and 2021. Finally, the desertification intensity map was calculated by geometric averaging for all three criteria for all three time periods. The results showed that 59.78% and 40.22% of the area of Tehran Province were in the low and medium classes, respectively. However, in 2016, the area of the medium class has increased to a 44.8%, and it is predicted that this increase will continue until 2021 so that 47.65% of the area of Tehran Province will be in the medium class. In addition, in this year, about 1% of the area of Tehran Province will be allocated to the high class in the western regions, which did not exist in the previous two periods. In general, due to human activities, the intensity of desertification in the western and southern parts of the province is higher than in the eastern and northern regions. [ABSTRACT FROM AUTHOR]

The aim of dynamic facility layout problem is to find the best layout for facilities at a multi period planning horizon so that the total cost of material handling and relocating the facilities is minimized. This paper developed a bi-objective mathematical model which is able to simultaneously minimize the material handling costs between facilities and the cost of rearrangement facilities and material handling time. Due to probabilistic characteristic of the transporters, such as the time of handling operation and existence of the failure, calculating the time required to carry the material using analytical relationships is impossible.Therefore, this paper uses the simulation approach and artificial neural networks. In this approach, a lot of scenarios are generated by combining of various levels of variables. Each scenario shows the location of the facilities and how transportation operations in each period is performed. Then each of these scenarios is implemented through computer simulation and simulation results are considered as the response variable. Finally, using input and response variable, an artificial neural network is trained to accurately estimate the time of carrying out the transportation operations. Given that the above problem is a NP-hard; this paper proposes a new meta-heuristic algorithm to optimize the problem and compares the performance of the proposed algorithm with existing algorithms in literature.

Unlike serial manipulators, forward kinematic problem (FKP) of parallel robots is highly complicated and its analytical solution is not generally le. Therefore, in most cases numerical methods are used to solve this problem which are relatively time consuming. On the other hand, reducing the duration of FKP analysis is an essential task for applications like real-time control. In this paper, artificial neural networks and a 3rd-order numerical algorithm are combined and a new hybrid strategy is proposed for forward kinematics analysis of parallel manipulators which significantly increases the speed of the FKP solution. In the proposed method, an approximate solution of the FKP is first produced by the neural network. This solution is next considered as an initial guess for the 3rd-order numerical technique which solves the nonlinear forward kinematics equations and obtains the answer with a desired level of accuracy. The proposed method is applied to a spatial 3-PSP parallel manipulator. The results show that using this method will lead to a 35 percent reduction in number of iterations and a 12 percent reduction in the FKP analysis time, while maintaining high level of solution accuracy. [ABSTRACT FROM AUTHOR]

Background: Warfarin is one of the most common oral anticoagulant, which role is to prevent the clots. The dose of this medicine is very important because changes can be dangerous for patients. Diagnosis is difficult for physicians because increase and decrease in use of warfarin is so dangerous for patients. Identifying the clinical and genetic features involved in determining dose could be useful to predict using data mining techniques. The aim of this paper is to provide a convenient way to select the clinical and genetic features to determine the dose of warfarin using artificial neural networks (ANN) and evaluate it in order to predict the dose patients. Methods: This experimental study, was investigate from April to May 2014 on 552 patients in Tehran Heart Center Hospital (THC) candidates for warfarin anticoagulant therapy within the international normalized ratio (INR) therapeutic target. Factors affecting the dose include clinical characteristics and genetic extracted, and different methods of feature selection based on genetic algorithm and particle swarm optimization (PSO) and evaluation function neural networks in MATLAB (MathWorks, MA, USA), were performed. Results: Between algorithms used, particle swarm optimization algorithm accuracy was more appropriate, for the mean square error (MSE), root mean square error (RMSE) and mean absolute error (MAE) were 0.0262, 0.1621 and 0.1164, respectively. Conclusion: In this article, the most important characteristics were identified using methods of feature selection and the stable dose had been predicted based on artificial neural networks. The output is acceptable and with less features, it is possible to achieve the prediction warfarin dose accurately. Since the prescribed dose for the patients is important, the output of the obtained model can be used as a decision support system. [ABSTRACT FROM AUTHOR]

In all societies and countries, in order to plan to provide the required energy for various sectors, it is necessary to accurately predict the demand, type of energy carriers and energy supply method. Considering the importance of food industries in each country, in this study, modeling of required energy for food industries sector was investigated. Modeling of energy consumption was performed using artificial neural networks. In the first step, the input data to the model was calculated according to statistics, balance sheets and input method proposed in this paper. Two methods, namely multiple neural network and single neural network were tested and the results showed that multiple neural network has a higher accuracy. For each of the energy carriers (gasoline, kerosene, fuel oil, natural gas, electricity, gasoline and LPG) the best neural network was selected by taking the average of 20 times per program for each network characteristic. Finally, the network was implemented in the form of final model using Simulink environment of MATLAB 7.0 software. Data analysis showed that daily consumption of natural gas in the industry is increasing, while the consumption of fuel oil and LPG is going to be decreased. [ABSTRACT FROM AUTHOR]

This paper presents the mathematical modeling and simulation of a cable-driven robotic device that can be used in gait rehabilitation of patients with lower extremity disabilities. A parallel cable robot is considered to assist a model of human body during walking. First, a proper pattern of walking is considered and kinematic and dynamic equations are solved to obtain tensions in cables for entire cycle of walking. By taking advantage of a numerical procedure, the workspace of the robot is investigated to find suspension points of the cables that would allow the model to remain in controllable workspace. Keeping the model in controllable workspace means that cables always remain in tension and robot can effectively engage in rehabilitation. The optimum locations are determined based on minimum cable tensions (energy consumption) and a Neural Network is trained to quickly determine suspension points based on anthropometric parameters of patient. The simulation results show the effectiveness of the method in tracking the desired trajectory of walking. The results of this study can be used for development and fabrication of an efficient cable driven rehabilitation system. [ABSTRACT FROM AUTHOR]

Considering the increasing rate of electrical energy usage, this energy has become one of the most important resources for human life. So all countries are seeking access to reliable and planned energy resource. Regarding the non-renewability of fossil fuel resources, especially oil and gas, the issue of replacing these types of energy with renewable energy has been considered for decades. Saving and optimal use of electrical energy in important applications such as residential and commercial buildings is critical. One of the most important factors for planning power consumption and optimizing it is accurate forecasting for next hours' power consumption of residential and commercial buildings. In this paper, first, the data sets of several residential buildings are analyzed using parallel wavelet converters. Then, using an optimal estimator model of the convolutional neural network, the short-term load of the building is estimated. The obtained results show that the proposed method has improved the prediction error about 70, 69 and 73 percent for ARIMA, SVR, and LSTM methods, respectively. [ABSTRACT FROM AUTHOR]

Given the special importance of customers in emerging organizations, this paper presents a customer loyalty model based on 8 factors influencing: customer loyalty including: trust, quality of service, transfer cost, reputation, commitment, perceived value, and satisfaction in 8 emerging private banks in Tehran. The data collection tool of this quantitative and survey strategy study was a questionnaire distributed among the statistical population of 340 members who were selected by random sampling. Statistical analysis was performed by the regression method and then the ANN model. The results of the regression model showed that the highest influence was for loyalty, trust, quality of service, and empathy on customer perceived value factors. And the most effect on customer satisfaction was for trust, quality of service, empathy, and commitment modes. Customer perceived value and satisfaction were directly related to customer loyalty. The results obtained from the artificial neural network model indicate that it has fewer errors than the regression model. [ABSTRACT FROM AUTHOR]

Objective One of the most important risks faced by financial institutions is credit risk, which is formed by the possibility of the receiver's default in repaying the loan or repaying the loan on time. One of the pivotal financial institutions in Iran's economy is the Export Guarantee Fund of Iran. Being the exclusive official export credit insurance entity in Iran, it operates under complete government ownership, shouldering all obligations associated with the fund's coverage. Over the past decade, the Export Guarantee Fund of Iran, the sole provider of export guarantees in the Iranian economy, has undertaken the risk amounting to approximately $15.8 billion through the issuance of diverse guarantees and insurance policies. The most significant volume of claims pertains to credit guarantees. Based on available statistics and official records within the Iranian entity, it can be observed that between the Iranian calendar years of 1387 and 1397, this fund issued guarantees totaling $6.7 billion, with the largest claim, exceeding $317 million, originating from the credit guarantees provided for the benefit of creditors. Recognizing the significance of this matter, the development of an optimal model for evaluating the credit guarantees provided by customers of the Export Guarantee Fund of Iran holds the potential to mitigate credit risk and subsequently decrease the fund's incurred losses. Even a one percent reduction in the fund's losses could result in substantial savings, amounting to at least 1,330 billion Rials. This figure is poised to grow even further as the fund increases the issuance of credit guarantees. Consequently, the primary objective of this research is to identify the most effective method for distinguishing between creditworthy and high-risk customers, ultimately diminishing the default rate of loans extended by the Export Guarantee Fund of Iran. Methods In this study, an artificial neural network model was employed to create a risk assessment framework for applicants seeking financial facilities and guarantees from this fund. This model exhibits exceptional predictive capabilities, especially in estimating the likelihood of default for the granted facilities. To achieve this objective, a dataset comprising 2,170 credit files granted by this fund was utilized. A total of 69 intra-organizational and extraorganizational variables were initially considered, and their respective information was gathered. Subsequently, six key variables were employed in the ultimate neural network model, encompassing the credit recipient's province, the guarantee beneficiary (creditor bank), personnel count, current ratio, real interest rate, and economic growth rate. Results Based on the neural network model, the best models were extracted with three methods: Bayesian, Levenberg, and Conjugate Gradient, and with one to 30 neurons in the hidden layer. The best model with all independent variables (i.e., 69 independent variables) was predicted with 96.2% accuracy, which is more than the Probit econometric model. Conclusion The findings of this research offer the capability to distinguish between customers with favorable and unfavorable credit profiles. Furthermore, the credit rating of customers can be used as a basis for adjusting the collateral requirements for each customer group, thereby aligning the collateral amount with the creditworthiness of the respective groups. This means that if the customer is one of the customers with good credit and a good credit rating, less collateral can be obtained from them, and if the customer has higher credit risk, more collateral can be used as a guarantee of the repayment obligation. In this case, while measuring the credit risk and separating customers, it is possible to manage the risk and the credit portfolio of the Export Guarantee Fund of Iran optimally. [ABSTRACT FROM AUTHOR]

Many efforts have been done in recent years to decrease the required time for analysis of FKP (Forward Kinematics Problem) of parallel robots.This paper starts with developing kinematics of a parallel robot and finishes with a suggested algorithm to solve forward kinematics of robots. In this paper, by combining the artificial neural networks and a 3rd-order numerical algorithm, an improved hybrid strategy is proposed in order to increase the accuracy and speed of forward kinematics analysis of parallel manipulators. First, an approximate solution of the forward kinematics problem is produced by the neural network. This approximate solution is then considered as the initial guess for the 3rd-order Newton-Raphson numerical technique. By applying Stewart-Gough parallel manipulator, the efficiency of the proposed method is evaluated. It is shown that replacing the Newton-Raphson algorithm by the 3rd-order one leads to a reduction of the iterations required to reach the desired accuracy level and thus a reduction of the FKP analysis time. Finally, Stewart robot is used to simulate the movement of jaw. This novel algorithm can be applied to any forward kinematics of serial or parallel robots.

Introduction: A total dissolved solid (TDS) is an important indicator for water quality assessment. Since the composition of mineral salts and discharge affects the TDS of water, it is important to understand the relationship of mineral salts composition with TDS. Materials and Methods: In this study, methods of artificial neural networks with Levenberg-Marquardt training algorithm, adaptive neuro fuzzy inference system based on Subtractive Clustering and Gene expression programming were used to model water quality properties of Zarrineh River Basin at upstream of Boukan dam, to be developed in total dissolved solids prediction. ANN and ANFIS programs code were written using MATLAB programming language. Here, the ANN with one hidden layer was used and the hidden nodes’ number was determined using trial and error. Different activation functions (logarithm sigmoid, tangent sigmoid and linear) were tried for the hidden and output nodes and the GeneXpro Tools 4.0 were used to obtain the equation of the best models. Therefore, water quality data from two hydrometer stations, namely Anyan and Safakhaneh hydrometer stations were used during the statistical period of 18 years (1389-1372). In this research, for selecting input variables to the data driven models the stepwise regression method was used. In the application, 75% of data set were used for training and the remaining, 25% of data set were used for testing, randomly. In this paper, three statistical evaluation criteria, correlation coefficient (R), the root mean square error (RMSE) and mean absolute error (MAE), were used to assess model’s performances. Results and Discussion: By applying stepwise method, the first significant (at 95% level) variable entered to the model was the HCO3. The second variable that entered to the model was Ca. The third and fourth ones were Na and Q respectively. Mg was finally entered to the model. The optimal ANN architecture used in this study consists of an input layer with five inputs, one hidden and output layer with three and two neurons for Anyan and Safakhaneh hydrometer stations, respectively. Similar ANN, ANFIS-SC5 model had the best performance. It is clear that the ANFIS with 0/4 and 0/7 radii value has the highest R and the lowest RMSE for Anyan and Safakhaneh hydrometer stations, respectively. Various GEP models have been developed using the input combinations similar ANN and ANFIS models. Comparing the GEP5 estimations with the measured data for the test stage demonstrates a high generalization capacity of the model, with relatively low error and high correlation. From the scatter plots it is obviously seen that the GEP5 predictions are closer to the corresponding measured TDS than other models. As seen from the best straight line equations (assume the equation as y=ax) in the scatter plots that the a coefficient for GEP5 is closer to 1 than other models. In addition to previous operation, Gene expression programming offered mathematical relationships in the stations of Anyan and Safakhane with the correlation coefficients, respectively 0.962 , 0.971 and with Root-mean-square errors, respectively 12.82 , 29.08 in order to predict dissolved solids (TDS) in the rivers located at upstream of the dam. The obtained results showed the efficiency of the applied models in simulating the nonlinear behavior of TDS variations in terms of performance indices. Overall, the GEP model outperformed the other models. For all of applied models, the best result was obtained by application of input combination (5) including HCO3, Ca, Na, Q and Mg. The results are also tested by using t test for verifying the robustness of the models at 95% significance level. Comparison results indicated that the poorest model in TDS simulation was ANN especially in test period. The observed relationship between residuals and model computed TDS values shows complete independence and random distribution. It is further supported by the respective correlations for GEP5 models (R2 = 0.0011 for Anyan station and R2 = 0.0123 for safakhaneh station) which are negligible small. Plots of the residuals versus model computed values can be more informative regarding model fitting to a data set. If the residuals appear to behave randomly it suggests that the model fits the data well. On the other hand, if non- random distribution is evident in the residuals, the model does not fit the data adequately. On the base of these results, we propose GEP, ANFIS-SC and ANN methods as effective tools for the computation of total dissolved solids in river water, respectively. Conclusion: It can be concluded that the ANN, ANFIS-GP, ANFIS-SC and GEP models can be considered as promising tools for forecasting TDS values, based on water quality parameters. It is notable from the results that the prediction accuracy of all applied models increases by increasing the number of input combinations. With attention to the aim of current research that is presenting the feasibility of artificial intelligence techniques for modeling TDS values, it is notable that the results presented in this paper are for research purpose and applying the abstained results for real-world needs some complicated steps and building artificial intelligences methods, based on complete data and parameters maybe affected the TDS values.

Electroencephalogram signals (EEGs) show the electrical activity of brain neurons. EEG is a non-invasive method that can be used to detect abnormal brain activities. Seizure is one of these abnormal activities and is the most common manifestation of epilepsy. Spikes are the most important characteristic of the seizure prone EEG signals. By detecting spikes, it is possible to detect epileptic seizures from EEG signals. EEG signals are non-stationary signals, so the wavelet transform that has appropriate time and frequency resolution can be a good option for extracting features of EEG signals. In this paper, after the extraction process using wavelet transform, artificial neural networks (ANNs) are used to classify healthy and epileptic signals. Particle swarm optimization (PSO) is also used as a novel approach to select weights and biases of network to improve network performance. The results of the implementation of the proposed algorithm have a 96.2% accuracy, which shows acceptable performance compared to existing methods. [ABSTRACT FROM AUTHOR]

Swelling percent is a very important physical property of the final product. Swelling test is timely and needed to cost. Therefore, swelling prediction of particleboard in production during can operate of the process line and consistent quality of production. In this research, variables such as moisture content of particle, the amount of adhesive, press time, press temperature, press pressure and swelling properties of particleboard were collected from Debalkhazae mill. The normalized data was analyzed by artificial neural network and fuzzy systems. Also, swelling percent was predicted by optimal model. The best model of swelling prediction is 5-5 basis of artificial neural networks, and the best function of fuzzy systems is Z-shaped curve membership function. Means absolute percent errors of the predictions are equal 5 and 22 percent, respectively. ANN method has better performance compared with fuzzy systems.

Nowadays, chain stores are experiencing a severe competitive environment. In such anenvironment, a growing number of customers are equipped by high volumes of informationand are more risk-taking. As they tend to switch their store based on the higher value thatthey might receive elsewhere, as customers switching behavior has become so commonplaceand has become a challenge for retailers. Thus, factors that influence the willingness ofcustomers to repurchase will play a critical role in the performance and success of chainstores. This paper aimed to identify and prioritize the determinative factors that causecustomers repeat patronage in Etka chain stores. Factors were extracted based on a thoroughliterature review and then these factors were modified using pre-test and interview methods.In this study, questionnaires were used for data collection. Respectively, to identify andprioritize determinative factors, correlation analysis and artificial neural networks were used.Results showed that in order of importance, quality, price loyalty program, number of servicecounters, store staff, shelf labels, product selection, parking facilities, home delivery, storebrand, and store atmosphere are factors that determine customers’ repeat patronage of Etkachain stores. Meanwhile, young, educated and wealthy customers showed lower store repeatpatronage. Finally, the paper concludes with suggestions for increasing consumer repeatpatronageof Etka chain-stores and directions for future research

Background and Aim Most heart diseases show symptoms on ECG, but diagnosing heart disease with ECG requires the knowledge and experience of medical specialized. Because these specialists may not always be available, it is necessary to design tools to diagnose heart disease in these situations. In this paper, a two-stage approach based on artificial neural networks is designed to diagnose heart disease using ECG information.In this study, we aim to propose a two-stage approach using artificial neural network (ANN) to diagnose heart disease based ECG data. Methods & Materials To design the proposed approach, first the ECG data of 861 patients referred to medical centers in Arak, Iran were collected. The data were examined based on the opinions of specialists. Then, 154 features from ECG were used as inputs to the proposed model. In the first stage, an ANN was used to detect the ECG status (usable and unusable). In the second stage, using the usable ECG data, an ANN was used to diagnose the presence or absence of heart disease. Finally, the performance of the two-stage approach was evaluated and its accuracy and precision in determining the ECG quality and heart disease diagnosis were determined. Ethical Considerations This study was approved by the ethics committee of Arak University of Medical Sciences (Code: IR.ARAKMU.REC.1400.138). Results In the proposed approach, the ANN used for the determining the ECG status had a precision of 97.1% and an accuracy of 97.3%. The ANN used for the diagnosis of heart disease had a precision of 95.8% and an accuracy of 95.4%. Conclusion Considering the high efficiency of the proposed approach in determining of ECG status and diagnosing heart disease, it is possible to use this approach to help the treatment staff.

The stability of frequency and voltage is one of the basic principles in the power systems. One of the latest control measures for power system frequency control and stability is load shedding. A fast and optimal adaptive load shedding method using Artificial Neural Networks (ANN) is presented in this paper. In this paper, the total power generation and the total existing load in power system were selected as the ANN inputs. This method has been tested on theNew England test system. The simulation results show the ability of this frequency control algorithm for optimal solving problem related to conventional method.

The Balochistan province of Pakistan is mostly affected by severe drought events due to little amount of precipitation. “Several Large Scale Climate Drivers (LSCDs) are known for their effects on precipitation worldwide but studies in the region are missing; a wide variety of LSCDs and lagged associative information”.The current study aimed to identify the significant LSCDs in the Balochistan province of Pakistan and improve the prediction skill of monthly precipitation by employing the Principal Component Analysis, Artificial Neural Network (ANN), Bayesian Regularization Neural Network (BRNN), and Multiple Regression (MR) Analysis using the 12 months lagged LSCDs such as Nino-1+2, Nino-3, Nino-3.4, Nino-4, QBO at 30 and 50hpa (QBOI and QBOII), Sea Surface Temperature (SST), 2m air temperature (T2M), 500hpa and 850hpa geopotential heights (H500 and H850), 500hpa zonal velocity (U500), 500hpa and 850hpa meridional velocity (V500 and V850), Latent and Sensible Heat Fluxes Over Land (LHFOL and SHFOL), and Surface Specific Humidity (SSH). To collect the data, Global Land Data Assimilation System, Tropical Rainfall Measuring Mission, MERRA-2, NOAA, Freie University Berlin, and HadISST datasets were used. The results of the study showed that significant LSCDs with a 99% confidence level were SSH, SST, LHFOL, SHFOL, T2M, U500, Nino-3.4, and Nino-4. During the test period, compared with MR models of 0.15 to 0.49 and principal components of -0.16 to 0.43, the ANN and BRNN models had better predictive skills with correlation coefficients of 0.40 to 0.74 and 0.34 to 0.70, respectively. It can be concluded that the ANN and BRNN models enable us to predict monthly precipitation in the Balochistan region with lagged LSCDs. Extended Abstract Introduction Pakistan is one of the most vulnerable countries due to climate change. According to the Global Climate Risk Index (GCRI) report, with the passage of time, its vulnerability is increasing. The main threats consist of a rise in temperature, irregular patterns of rainfall, a rise in the sea level, and extreme events such as droughts, floods, and heatwaves. Precipitation plays an important role in the economy of Pakistan as an agricultural country that has experienced variations, particularly in a recent couple of decades when a sharp jump in global atmospheric temperatures was noticed. The southern part of Pakistan which comprises the Sindh and Balochistan provinces has an arid climate and is mostly affected by severe drought events due to less amount of rainfall throughout the year as compared to other parts of the country. This study’s main focus is Balochistan province which has a long history of severe droughts and is an important province in terms of agriculture. Less than 250mm is the annual average rainfall received by the region. It is predicted that throughout the 21st century if adaptation measures are not taken, there will be a continuous increase in droughts and scarcity of water which will adversely affect the lives of people and the economy. Globally, precipitation and droughts are strongly linked with the large-scale climate drivers through atmospheric associations but there are no detailed studies in Balochistan province that consider a wide variety of large-scale climate drivers and their lagged association with precipitation. The main objective of the present study is to identify the significant large-scale climate drivers in the Balochistan province of Pakistan and improve the prediction skill of monthly precipitation by applying the principal component analysis, artificial neural network, Bayesian regularization neural network, and multiple regression analysis considering the lagged association of climate drivers. Methodology In the present study, first, large-scale climate drivers from NOAA HadISST, MERRA-2, and Freie University Berlin in NetCDF files were processed using the ArcGIS model builder along with the precipitation data from GLDAS and TRMM. The data were first normalized in the range of 0 and 1 using the min-max normalization formula. For showing the 12 months’ lagged association between the large-scale climate drivers and precipitation, a cross-correlation method was employed and the heatmaps were created in Origin 2021b software to show significant lagged correlations. Principal component analysis was applied and the variances and Eigenvalues for each of the components were calculated. Multi-layer feed-forward neural network with a back propagation algorithm was used for the prediction of monthly precipitation with the first two PCs in most cases, three and one in some cases as the predictors. Bayesian Regularization Neural Network (BRNN) was applied which is a version of Artificial Neural Networks (ANN) and is a more powerful method as compared to conventional ANN. MATLAB R2015a environment was used for this purpose. Then, Multiple Regression Analysis (MRA) was carried out and was used as a benchmark for comparing ANN and BRNN models using significantly lagged climate indices in one case and in another case the selected PCs. To validate the performance of all developed models, the TRMM dataset was used. Finally, time series graphs and Radar charts were prepared for comparison. Discussions The findings of cross-correlations between monthly precipitation and large-scale climate drivers in Balochistan province showed the most dominant climate drivers as surface specific humidity, sea surface temperature, latent and sensible heat fluxes over land, 500hpa zonal velocity, 2m air temperature, Nino-3.4 and Nino-4, and the highest correlations were noticed for surface specific humidity, sea surface temperature, latent and sensible heat fluxes over land, 500hpa zonal velocity, 2m air temperature. It is noteworthy that in each district, the lagged correlation of maximum climate indices with precipitation was distinct at distinct lags. The multiple regression best models were selected on the basis of no violation of the limits of statistical significance and lower errors. The MLR model’s performance was low in the region where some of the districts had low correlations. The results from ANN and BRNN models showed that the BRNN models were on a lower side than the ANN models with higher values of correlation coefficient showing their capability of finding the pattern and trend of the GLDAS precipitation. The results of the evaluation of the generalization capability of all models on the TRMM dataset showed that both ANN and BRNN models relatively performed well as correlation and error values were closer to each other. Conclusions The results of this study showed that, first, significant LSCDs with 99% confidence level were surface specific humidity (SSH), sea surface temperature (SST), latent and sensible heat fluxes over land (LHFOL and SHFOL), 2m air temperature (T2M), 500hpa zonal velocity (U500), Nino-3.4, and Nino-4. Second, to predict the monthly precipitation using lagged LSCDs and principal components, MR models were developed. MR models’ performance was low. The highest Pearson correlation in MR models during the training set was observed for Musakhel district as 0.65 and during the test as 0.49 for Musakhel and Zhob. In MR-PC models, the highest correlation was recorded for Awaran (0.50) during the training and for Quetta (0.43) during the test. Third, ANN and BRNN models were developed using the selected PC components and gave higher correlations as compared to regression models indicating their capability of finding the pattern and trend of the observations. They generally manifested lower errors and are more reliable for the purpose of prediction in the region. Maximum correlations during training were 0.77 and 0.73 in both ANN and BRNN models, respectively, for the Lasbela district. In the test case, they were 0.74 and 0.70 for Awaran. Their generalization ability was tested on the TRMM dataset. In conclusion, this study divulged the possibility of monthly precipitation prediction using ANN and BRNN and lagged LSCDs for the study region. It is explicit that the response of precipitation to climatic factors is delayed. Comparing the results of the ANN, BRNN, and MLR, it can be concluded that the artificial intelligence approaches such as ANN and BRNN are reliable nonlinear statistical options that can generate similar and in some cases better forecasts and could be useful for the agriculture and water management in the Balochistan region as it is a very important part of Pakistan in terms of agriculture. In the future, this research will be extended by developing the prediction models by including more LSCDs and can be improved by the approaches of genetic algorithm for the best input selection and the use of observed station-based precipitation data sets. [ABSTRACT FROM AUTHOR]

• One of the most challenging safety problems in open pit mines is backbreak during blasting operation, and its prediction is very important for a technically and economically successful mining operation. To avoid backbreak, different parameters such as physicomechanical properties of rock mass, explosives properties and geometrical features of the blasting pattern should be considered. This paper presents a new solution of multiple linear regression (MLR), particle swarm optimization algorithm (PSO) and artificial neural networks (ANNs) to estimate the backbreak induced by bench blasting, based on major controllable blasting parameters. To this aim, Angouran mine in Iran was considered and blasting pattern parameters for 73 operations were collected. In addition, back-break was measured in each operation. Considering the previous investigations and also collected data from the mine, burden, spacing, hole length, stemming, charge per delay, RQD, number of row and powder factor were selected as input parameters. In order to find the better solutions, the constructed models were implemented in PSO algorithms. Also, the prediction of backbreak was investigated using ANNs. According to the obtained results, the PSO algorithm is a suitable tool for optimizing models and obtaining more accurate prediction of backbreak. Among the presented empirical models, the optimized exponential model with PSO algorithm with a RMSE (0.31) and R2 (0.87) shows the better results in prediction of backbreak and it is suitable for practical use in Angouran mine. Considering the sensitivity analysis, among the input parameters, length of stemming and charge per delay have shown the most and the least effect on the backbreak, respectively. The results of ANNs show that multilayer networks are more powerful and efficient than single-layer in prediction of backbreak.

Application of non-classical methods in modeling complex systems and forecasting their behavior has become as more as usual for the scientists and professionals. In most complex systems, especially in non-linear systems, application of classical methods is very difficult or even useless. Non-classical methods are intelligent, knowledge-based, very flexible, and therefore effective in modeling and forecasting. Neural Networks are one of the well-known and innovative nonclassical methods, which have being used in modeling, pattern recognition, clustering and forecasting. This paper tries to predict the economic time series by neural nets. Economic time series are considered as outputs of complex and non-linear economic systems, which can be modeled and forecasted by the neural nets.It has been shown that the performance of neural nets (as prediction machine) is very sensitive and dependent on the structure, size, and learning method of neural nets. In this paper, using MATLAB neural nets toolbox, some Iranian economic time series are being used as case studies for the neural network application.

Lack of accurate knowledge of pavement behavior under moving loads is the one of the most important disadvantages in calculation of Equivalent Axle Load Factor (EALF) in roads pavement. Among the many researches, the most comprehensive method is based on the AASHTO road test. As the main weakness of this method, the results are limited to the experimented axles, which makes it impossible to determine the EALF for all existing axles, hence reducing the accuracy of the results, causing premature failure, and leading to higher maintenance costs. Today, although numerous software packages are available for EALF calculation, they require various parameters, are time-consuming, and can only simulate one section at a time . On the other hand, artificial neural networks, as an artificial intelligence subcategory, have many advantages such as reduced input data, increased modeling process speed, ability of parallel modeling of several pavements with different conditions, etc. In this paper, after verifying the simulation of flexible pavements in ABAQUS, a model based on Artificial Neural Network (ANN) was presented to calculate EALF using the back-propagation architecture. Finally, from among the reviewed ANN configurations, the network with the 7-13-1 architecture incorporating the sigmoid function was selected as the optimum network.

Objective: With the advent of the Internet as the main channel for the supply of products and services in recent years, customers' experience of online shopping has become vital. The purpose of this research was segmenting customer's experience of online retail stores. Methodology: This study was an applied research in terms of purpose, and a descriptive-survey one in terms of method. The statistical population included all the people, who had an experience of buying from online retailers. The required sample size was calculated by using Cochran's formula for infinite communities. From among them, 384 people were selected through the available nonprobability sampling method. The data collection tool was a standard questionnaire, the validity and reliability of which were confirmed by content validity and Cronbach's alpha coefficient, respectively. To analyze the data and segment the customers, the approach of Self-Organizing Maps (SOM) was taken based on Artificial Neural Networks (ANNs) by using ViscoverySOMine software. Results and Discussion: According to the findings, the customers, who had an experience with online retailers, were divided into 3 segments with different demographic characteristics and components of online experience. The three categories of customers were apathetic, utilitarian, and visual customers. The first segment (apathetic customers) included those who were least affected by the components of the experience. The ability to communicate with the seller was most important to this group. The second group (utilitarian customers) consisted of the young and low-income women, who had the least online shopping in the short time. Benefit and trust were more important to the customers of this group. The young to middle-aged middle-income women constituted the third group (visual customers). These women often had online shopping for a long time. These people attached great importance to the product availability and paid more attention to store familiarity compared to other two clusters. [ABSTRACT FROM AUTHOR]

Handwriting Analysis has been used for a very long time to analyze an individual’s suitability for a job, and is in recent times, gaining popularity as a valid means of a person’s evaluation. Extensive Research has been done in the field of determining the Personality Traits of a person through handwriting. We intend to analyze an individual’s personality by breaking it down into the Big Five Personality Traits using their handwriting samples. We present a dataset that links personality traits to the handwriting features. We then propose our algorithm - consisting of one ANN based model and PersonaNet, a CNN based model. The paper evaluates our algorithm’s performance with baseline machine learning models on our dataset. Testing our novel architecture on this dataset, we compare our algorithm based on various metrics, and show that our novel algorithm performs better than the baseline Machine Learning models.

Financial crises in banking systems are due to inability to manage credit risks. Credit scoring is one of the risk management techniques that analyze the borrower's risk. In this paper, using the advantages of computational intelligence as well as soft computing methods, a new hybrid approach is proposed in order to improve credit risk management. In the proposed method, for modeling in uncertainty conditions, parameters of the neural network, including weights and errors, are considered in the form of fuzzy numbers. In this method, the underlying system is firstly modeled using neural networks and then, using fuzzy inferences, the optimal decision will be determined with the highest degree of superiority. Empirical results of using the proposed method indicate the efficiency and high accuracy of this method in analyzing credit rating problems.

Prediction of satellite orbital position is a critical requirement for all satellite ground stations. In this paper, a new viewpoint for predicting orbital position of satellites is presented. In contrast to traditional methods which are based on Kepler's law, the proposed method, is solely based on past observation of a given satellite. In contrast to traditional algorithms which have neglected some of the orbital perturbations, the most important feature of this method is considering all orbital perturbations by using real data. TLEs (Two Line Element sets) are the most available real data and are used in this research as the main data source. Using the capability of neural networks for time series prediction over available data, results in a fast and accurate orbital position predictor. The comparison between the output of our proposed method, SPG4 (Simplified General Perturbation version 4) propagator and real orbital position of a given satellite, shows the effectiveness of this algorithm.

Stock price forecasting is one of the main challenges in stock market which investors and analysts are faced with. To forecast the future prices and future trend, different tools have been used among which we can refer to technical and fundamental analysis. It is noticed that technical analysis has good performance in short-time forecasting. Hence, in this paper, technical analysis has been used to estimate the probability function of stock prices. To forecast the direction of stock price movement in the following day, artificial neural networks (ANN), Logit, Probit, and extreme value models are utilized. To evaluate the performance of proposed models, daily values of Iran Khodro company stock are considered as a real case study. The nonparametric test of equality of ratios shows that the difference between the forecasting results of different models is not statistically significant. However, according to forecasting error criterion, the Probit model is more efficient than other mentioned models.

Today most developed and developing countries emphasize on the importance of productivity as one of the necessities of economic development and competitiveness in the world. Because todays, competition is taking various dimensions and striving for higher productivity is one of the important factor of these competitions. On this basis, identifying the factors of affecting productivity growth in the Iran economy is essential for economic growth and development. Therefore, this study intends to first identify the factors affecting productivity growth by using feature selection logic, basis on Non-Dominated Sorting Genetic Algorithm (NSGA-II) then estimate the selective model using Artificial Neural Networks (ANN) for the period (1991-2016) and finally using the Garsen index to measure the sensitivity analysis of factors affecting productivity growth. Based on the results of the feature selection among the twenty variables, foreign investment, health investment, rail lines, innovation index and exchange rate (five variables) were removed from the model. Based on the results of ANN model with Tansig activation function with 3 neurons, it has a prediction power of 0.993 and minimum error of model 0.0019. Also, according to the Garsen index, human capital (15%), government size (11%), openness, research and development and economic corruption control (8%) had the highest impact on productivity growth and monetary development (1.48%) the rule of law (2.27%) and physical capital (3.2%) had the least impact on productivity growth. [ABSTRACT FROM AUTHOR]

Nowadays, many researches have been done to improve the efficiency of the nuclear power plants, one of which is the use of the dual cooled annular fuel which is an internally and externally cooled annular fuel with many advantages in heat transfer characteristics. Also, some studies have suggested that the usage of the nanoparticles in a fluid as nano-fluid can provide dramatic improvements in the thermal properties of fluid. Howere, the usage of neutronics and the thermal hydraulic codes in order to investigate the nano-fluid effects in the nuclear reactors is complex, uneconomical and time consuming. Therefore, in this paper, to investigate the nano-fluid effects on the important parameters of the VVER-1000 nuclear reactor with dual-cooled annular fuel, effects of Al2O3/ water nano-fluid concentration and size on neutronics and the thermal-hydraulic parameters in the VVER-1000 nuclear reactor are investigated using a proper Artificial Neural Network. Results show that the trained Neural Network has good convergence and accuracy in determination of the neutronics and the thermal-hydraulic parameters. Using the presented Neural Network, important reactor parameters can be determined without neutronics and the thermal hydraulic codes, thus saving time.

Background: One of the common diseases is fatty liver disease, which can be seen significantly in patients with excessive obesity as well as patients with diabetes. Using ultrasound imaging methods, it can be possible to identify and evaluate patients affected by this disease. The aim of this study was to use advanced machine learning algorithms for better and more accurate classification of the acquired images from different patients affected by fatty liver disease. Materials and Methods: In this study, the acquired ultrasound images of 55 patients suspicious as having fatty liver disease have been used. The level of fat for different patients was labeled by biopsy sampling. Based on this experiment, the patients were predicated as having a fatty liver when more than 5% of their liver hepatocytes were accompanied by infiltration of fat. Then, we utilized some pre-trained convolutional neural networks, including Inception-ResNetV2, GoogleNet, and AlexNet to extract high-level features of the collection of the acquired images. After that, a SoftMax layer was implemented to classify the images that indicated fatty liver. Results: The resulted precisions for Inception-ResNetV2, GoogleNet, and AlexNet pre-trained convolutional networks were 0.8108, 0.9459, and 0.9932, respectively. Also, the area under the curve of receiver operating characteristics (AUC) for these networks were 0.9757, 0.9960, and 0.9963, respectively. Conclusion: The proposed intelligent algorithm can help sonography experts to recognize the liver tissues with fat automatically and accurately without the need for a specialist to assign the region of interest for evaluation. [ABSTRACT FROM AUTHOR]

In this study, a different approach to the prediction of satellite position is introduced.All methods are based on the Kepler’s laws of planetary motion and the orbitalperturbations such as the Earth’s oblateness, atmospheric drag, third-body perturbationand the solar-radiation pressure. All these perturbations are modeled and are includedseparately in the equation. However, this paper offers a new view of the prediction whichsuggests the use of artificial neural networks and observation data. The advantage of thismethod is based on the usage of observation data, so that all disturbances are taken intoaccount and there is no need to use perturbation models. For this reason, the use of theTLE as the most reachable actual data is considered. Comparison of the output of thismethod with actual data shows the accuracy of the proposed method which is very high.

Geo-statistical methods for reserve estimation are difficult to use when stationary conditions are not satisfied. Artificial Neural Networks (ANNs) provide an alternative to geo-statistical techniques while considerably reducing the processing time required for development and application. In this paper the ANNs was applied to the Choghart iron ore deposit in Yazd province of Iran. Initially, an optimum Multi Layer Perceptron (MLP) was constructed to estimate the Fe grade within orebody using the whole ore data of the deposit. Sensitivity analysis was applied for a number of hidden layers and neurons, different types of activation functions and learning rules. Optimal architectures for iron grade estimation were 3-20-10-1. In order to improve the network performance, the deposit was divided into four homogenous zones. Subsequently, all sensitivity analyses were carried out on each zone. Finally, a different optimum network was trained and Fe was estimated separately for each zone. Comparison of correlation coefficient (R) and least mean squared error (MSE) showed that the ANNs performed on four homogenous zones were far better than the nets applied to the overall ore body. Therefore, these optimized neural networks were used to estimate the distribution of iron grades and the iron resource in Choghart deposit. As a result of applying ANNs, the tonnage of ore for Choghart deposit is approximately estimated at 135.8 million tones with average grade of Fe at 56.14 percent. Results of reserve estimation using ANNs showed a good agreement with the geo-statistical methods applied to this ore body in another work.

It is very important for investors and creditors to understand the critical factors affecting a firm’s value before making decisions about investments and loans. Since the knowledge- based economy has evolved, the method for creating firm value has transferred from traditional physical assets to intangible knowledge. Therefore, valuation of intangible assets will become a widespread topic of interest in the future of the economy. This paper is aimed, using the artificial neural networks, to identify the critical factors affecting intangible assets value in firms listed in Tehran Stock Exchange. The results show the importance of firm’s attributes, corporate governance, ownership structure and intangible capital in determining the value of intangible assets, respectively.

Soil erosion is undoubtedly one of the most important problems in natural areas of Iran and has destructive effects on different ecosystems. Considering that calculation of the sediment rate in sediment stations and direct measurements of erosion process is costly and difficult, it is critical to find ways to accurately estimate the amount of sediment yield in catchments especially in arid and hyper arid areas because of their high ecological sensitivity. One of the most commonly used methods in these areas is the sediment rating regression method. Therefore, in this study sediment observed data for 48 events (the corresponding discharge and sediment) in a 23-year period from Fkhrabad basin (Mehriz) were compared to the estimated data obtained from Multi-line rating method, extent middle class, middle class rating curve with correction factor QMLE, SMEARING correction coefficient FAO and Artificial Neural networks (ANNs). Finally, the accuracy of these methods were assessed using different evaluation criteria such as Root Mean Square Error (RMSE), coefficient of determination (R2) and the standard Nash (ME). Results showed that ANN outperformed the other methods with the RMSE, R2 and ME of 203.3, 0.86 and 0.66, respectively. The results suggest that these methods should be used cautiously in estimating the suspended sediment load in arid and hyper arid regions due to the nature of the observed data and temporal and seasonal flow systems in these regions. It was also indicated that the artificial neural network models have higher flexibility than other methods which makes them to be useful tools for modeling in poor data conditions. [ABSTRACT FROM AUTHOR]

Estimation of evaporation values is needed for efficient management of water resources at semi-arid regions. This paper presents application of Artificial Neural Networks (ANNs), Multiple Linear Regression (MLR) and empirical models viz.: Energy balance ، Aerodynamic ، Penman for estimation of daily pan evaporation for Tabriz and Urmia cities. Furthermore, in order to determine the effect of each input parameter on the output variable in terms of magnitude and direction and also identify the best combinations of the model inputs, two sensitivity analysis methods i.e. the Partial Derivation method (PaD) and the Weights method have been applied on the ANNs results. The used hydrological variables include daily observations of air temperature, pan evaporation, solar radiation, air pressure, relative humidity, and wind speed. The results of the classic methods and ANN models are compared to daily observations of evaporation values. The comparison showed that there is better agreement between the ANN estimations and measurements of daily evaporation than other models. Sensitivity analysis results showed that air temperature, solar radiation and previous day evaporation have maximum effects on daily evaporation in both regions and the contributions of the other variables are insignificant.