Search

With the development of computer technology, automatic test paper generation systems have gradually become an effective tool for detecting and maintaining national machine security and protecting the rights and interests of workers. This article achieved multi-level oral scores for different types of questions through online scoring using artificial neural networks in recent years. Based on its specific situation and evaluation index requirements, an analysis module that is reasonable, efficient, and in line with the hierarchical structure and module requirements of national conditions has been designed to complete the research on automatic test paper generation technology, in order to help better manage and allocate human resources and improve production efficiency. Afterwards, this article conducted functional testing on the technical module. The test results showed that the scalability of the system was over 82%. [ABSTRACT FROM AUTHOR]

To enhance solution accuracy and training efficiency in neural network approximation to partial differential equations, partitioned neural networks can be used as a solution surrogate instead of a single large and deep neural network defined on the whole problem domain. In such a partitioned neural network approach, suitable interface conditions or subdomain boundary conditions are combined to obtain a convergent approximate solution. However, there has been no rigorous study on the convergence and parallel computing enhancement on the partitioned neural network approach. In this paper, iterative algorithms are proposed to enhance parallel computation performance in the partitioned neural network approximation. Our iterative algorithms are based on classical additive Schwarz domain decomposition methods. For the proposed iterative algorithms, their convergence is analyzed under an error assumption on the local and coarse neural network solutions. Numerical results are also included to show the performance of the proposed iterative algorithms. [ABSTRACT FROM AUTHOR]

This paper discusses the estimation of zonal geomagnetic indices of two super geomagnetic activities of 2015 in the 24th solar cycle. It estimates the zonal geomagnetic indices (Dst, ap, AE) of 17 March and 22 June 2015 super storms with an artificial neural network model. The activities that happened in March and June are considered on the solar wind parameters (B z , E, P, N, v, T) and zonal geomagnetic indices obtained from NASA coordinated data analysis web. Descriptive values of the variables are indicated, binary correlations of the data are shown with the covariance matrix and the hierarchical cluster of the data are presented by the dendrogram. In the paper, the physical principles govern the artificial neural network model. The model utilizes solar wind parameters as inputs and zonal geomagnetic indices as outputs. The causality principle shapes the models by cause-effect relationship. Back propagation algorithm is specified as Levenberg–Marquardt (trainlm) and 30 neural numbers are used in the artificial neural network. The neural network model estimates the Dst, ap, and AE indices of 17 March and 22 June activities with reliable accuracy. The geomagnetic activity estimation can support interplanetary studies. [ABSTRACT FROM AUTHOR]

As the deployment of artificial intelligence (AI) algorithms at edge devices becomes increasingly prevalent, enhancing the robustness and reliability of autonomous AI-based perception and decision systems is becoming as relevant as precision and performance, especially in applications areas considered safety-critical such as autonomous driving and aerospace. This paper delves into the robustness assessment in embedded Deep Neural Networks (DNNs), particularly focusing on the impact of parameter perturbations produced by single event upsets (SEUs) on convolutional neural networks (CNN) for image semantic segmentation. By scrutinizing the layer-by-layer and bit-by-bit sensitivity of various encoder–decoder models to soft errors, this study thoroughly investigates the vulnerability of segmentation DNNs to SEUs and evaluates the consequences of techniques like model pruning and parameter quantization on the robustness of compressed models aimed at embedded implementations. The findings offer valuable insights into the mechanisms underlying SEU-induced failures that allow for evaluating the robustness of DNNs once trained in advance. Moreover, based on the collected data, we propose a set of practical lightweight error mitigation techniques with no memory or computational cost suitable for resource-constrained deployments. The code used to perform the fault injection (FI) campaign is available at https://github.com/jonGuti13/TensorFI2 , while the code to implement proposed techniques is available at https://github.com/jonGuti13/parameterProtection. [ABSTRACT FROM AUTHOR]

In recent years, the application of deep neural networks in the field of image classification has become increasingly widespread. However, existing researches have exposed the inherent vulnerabilities and opacity of these network models, rendering them susceptible to carefully crafted adversarial attacks. This poses significant security risks to their practical deployment. Given that the majority of models in real-world applications operate in a black-box manner, current research on black-box attacks targeting classification models remains inadequate, facing significant challenges such as incomplete robustness evaluations, low success rates of black-box attacks, and excessive resource consumption during the attack process. To address these challenges, this paper innovatively proposes a black-box attack algorithm based on a cycle-consistent generative network. This algorithm directly generates adversarial perturbations through a generative network that integrates an attention mechanism, deeply learning the complex relationship between adversarial samples and clean sample data. It supports both targeted and untargeted attack modes. Experimental validation on the SAT dataset demonstrates that our method achieves an average white-box attack success rate exceeding 96% in classification tasks and an average success rate of up to 69.4% in black-box transfer attacks, exhibiting excellent transferability across different models. This research contributes significantly to enhancing the effectiveness of black-box attacks, reducing attack costs, and providing novel ideas and methods for enhancing the security of deep learning models. • We proposed an optimized cycle-consistent generative network for adversarial attacks, ultimately enhancing the success rate against other diverse black-box models. • Our algorithm offers support for both targeted and non-targeted attacks. Our targeted approach confers directionality to the generated perturbations, enabling purposeful attacks tailored to the attacker's intentions. • A transformation distribution module is incorporated into our model. We introduce a distance loss function to our overall loss computation. • Our approach enables migration attacks across different models, achieving a black-box attack success rate superior to current methods, exceeding 69.4%. [ABSTRACT FROM AUTHOR]

Abstract In the last 20 years, and in particular in the last decade, the availability of propagation data for GNSS has increased substantially. In this sense, the ionosphere has been sounded with a large number of receivers that provide an enormous amount of ionospheric data. Moreover, the maturity of the models has also been increased in the same period of time. As an example, IGS has ionospheric maps from GNSS data back to 1998, which would allow for the correlation of these data with other quantities relevant for the user and space weather (such as Solar Flux and Kp). These large datasets would account for almost half a billion points to be analyzed. With the advent and explosion of Big Data algorithms to analyze large databases and find correlations with different kinds of data, and the availability of open source code libraries (for example, the TensorFlow libraries from Google that are used in this paper), the possibility of merging these two worlds has been widely opened. In this paper, a proof of concept for a single frequency correction algorithm based in GNSS GIM vTEC and Fully Connected Neural Networks is provided. Different Neural Network architectures have been tested, including shallow (one hidden layer) and deep (up to five hidden layers) Neural Network models. The error in training data of such models ranges from 50% to 1% depending on the architecture used. Moreover, it is shown that by adjusting a Neural Network with data from 2005 to 2009 but tested with data from 2016 to 2017, Neural Network models could be suitable for the forecast of vTEC for single frequency users. The results indicate that this kind of model can be used in combination with the Galileo Signal-in-Space (SiS) NeQuick G parameters. This combination provides a broadcast model with equivalent performances to NeQuick G and better than GPS ICA for the years 2016 and 2017, showing a 3D position Root Mean Squared (RMS) error of approximately 2 m. [ABSTRACT FROM AUTHOR]

Today, the Structural Health Monitorin (SHM) methodology is the main way to deal with the detection and identification of damages in a range of engineering sectors, mainly in the military and civil aerospace industry. The need to monitor damages and failures in increasingly complex structures has led to the development of various detection techniques. Identification of damages through intelligent signal processing and optimization algorithms are particularly emphasized. The methods discussed here are mainly elaborated by the evaluation of modal data due to the great potential of their application. Moreover, the optimization of damage identification is approached through methods of optimal positioning of sensors for the acquisition of the data that must be evaluated so that conclusions can be made. This article discusses the use of computational and intelligent techniques for structural monitoring in the form of a review with emphasis on composite materials. Despite the excellent mechanical performance already known about composite materials, they have a weak point. While damages, in a metallic material are easily visible (in some cases), composite materials often have the superficial appearance as if in perfect condition, when, inside, there are serious damages. This paper can be seen as a guideline or a starting point for developing and improving SHM systems. The contents of this paper aim to help engineers and researchers find a starting point in developing a better solution to their specific structural monitoring problems, either by inverse methods, pattern recognition, and intelligent signal processing. [ABSTRACT FROM AUTHOR]

The primary purpose of this paper is to realize robust place recognition algorithms towards simultaneous viewpoint and condition changes, and provide satisfactory computational efficiency. In this paper, we significantly improve the viewpoint invariance of the SeqSLAM algorithm by using state-of-the-art deep learning techniques to generate robust feature representations of images and develop the SeqCNNSLAM . Experimental results show that SeqCNNSLAM outperforms state-of-the-art place recognition systems in most cases, such as, when precision is maintained at 100%, the maximum recall obtained by SeqCNNSLAM is 50% higher than SeqSLAM on the Norland dataset with simultaneous condition change and 12.5% viewpoint change. Besides, we develop an acceleration method called A-SeqCNNSLAM , which exploits the location relationship between the matching images of adjacent images to reduce the matching range of the current image. Experimental results demonstrate that an acceleration of  ∼  5 times is achieved with minimal accuracy degradation of  ∼  5%. Finally, to enable A-SeqCNNSLAM adaptability in new environments, O-SeqCNNSLAM is devised for the online parameter adjustment in A-SeqCNNSLAM. [ABSTRACT FROM AUTHOR]

In the artificial neural networks domain, the Levenberg-Marquardt technique is novel with convergent stability and generates a numerical solution of the wire coating system for Sisko fluid flow (WCS-SFF) through regression plots, histogram representations, state transition measures, and means squared errors. In this paper, the analysis of fluid flow problem based on WCS-SFF is studied with a new application of intelligent computing system via supervised learning mechanism using the efficacy of neural networks trained by Levenberg-Marquardt algorithm (NN-TLMA). The original mathematical formulation in terms of PDEs for WCS-SFF is converted into dimensionless nonlinear ODEs. The data collection for the projected NN-TLMA is produced for parameters associated with the system model WCS-SFF influencing the velocity using the explicit Runge-Kutta technique. The training, validation, and testing processes of NN-TLMA are utilized to evaluate the obtained results of WCS-SFF for various cases, and a comparison of the obtained results is performed with reference data set to check the accuracy and effectiveness of the proposed algorithm NN-TLMA for the analysis of non-Newtonian fluid problem-related WCS-SFF. The proposed NN-TLMA for solving the WCS-SFF is effectively confirmed through state transition dynamics, mean square error, regression analyses, and error histogram studies. The powerful consistency of suggested outcomes with reference solutions indicates the validity of the framework, and the accuracy of 10 - 8 to 10 - 6 is also achieved. [ABSTRACT FROM AUTHOR]

The paper presents an approach to machine learning involving the use of a hybrid method in Neural ODE. The authors consider a way of the Neural ODE application with an embedded hybrid method for intelligent manipulator control in a non-deterministic environment. To reduce computational resources, the number of layers of the neural network using ODE is reduced, the solution of which is carried out by a hybrid transformation method. The hybrid method includes normalizing the system, linear transformation, and polynomial transformation. Using the hybrid method in Neural ODE can increase the learning and computation speed. The transformation method has high accuracy with low computational complexity of the algorithm. As a result of applying the hybrid method, the original nonlinear system is transformed to a simpler form. To solve the simplified transformed system, the scheme of the numerical Runge-Kutta method of the third order with an adaptive step is used. The results of the computational experiment show an increase in the speed and accuracy of the calculation. The presented approach to implementing the hybrid method in Neural ODE allows faster training of the neural network of the world model on limited data obtained from the real environment. After training on the data generated by the model of the world, a robotic arm always successfully reaches the object during subsequent interaction with the real environment. This solves the problem of the limited ability of the agent to interact when learning with the real environment in robotic systems. [ABSTRACT FROM AUTHOR]

In this study, Artificial Neural Networks (ANN) were used to model parameters of scheduling of reconfigurable embedded systems containing resource constraints for applications running in real-time. The main goal is to implement a neural network based approach for real-time scheduling in order to handle real-time constraints in execution scenarios. Many techniques have been proposed for both the planning of tasks and reducing energy consumption. This paper presents a new hybrid contribution that handles the real-time scheduling of embedded systems by keeping energy consumption at a low power depending on the combination of Dynamic Voltage Scaling (DVS) and the energy Priority Earlier Deadline First (PEDF) algorithm. Indeed, in our original proposed approach, an other combination of DVS and time feedback can be used to scale the frequency by dynamically adjusting the operating voltage. The originality of our algorithm appears by allowing medium priority tasks to be executed more quickly before their deadline while decreasing their ability to be send again to the waiting list in order to ensure the execution of a task with the lowest voltage possible. [ABSTRACT FROM AUTHOR]

Temperature variations, climatic changes or human interventions are the main cause buildings and streets deteriorate over time. Since manual verification of such structures completely depends on a specialist's knowledge and experience it is often time-consuming, costly, and dangerous, therefore an automated approach would be very beneficial. Automated devices using computer vision that detect anomalies in targeted structures have the potential of creating a better and safer environment, periodically checking for possible faults and improving worker safety. The system presented in this paper is an Android application with the purpose of field use, the goal being that either the worker could use it to detect potential hazards or an autonomous system might use it in inhospitable scenarios. The model created contains two modules, the U-Net algorithm, and the Android application; the novelty of the approach comes from offline use and accuracy, optimizations have been made in order for a reasonably powerful device to run the model and classify in real-time without the use of servers or mobile data. Adapting the neural network model to the mobile application was a challenge for our team, but in the end, the mobile application works as we intended with good results. [ABSTRACT FROM AUTHOR]

Drug names recognition (DNR) that seeks to discriminate drug mentions within the unstructured biomedical texts and categorize them into drug classes, is an essential task in mining medical information. The neural network structure considerably influences their overall performance. If there were lots of neurons within a network, it might fit the training dataset better, and the network might achieve few training errors, however it could be prone to over-fitting and the poor network generalization. Therefore, this paper proposes a novel network model for DNR based on a hybrid genetic artificial bee colony (GABC) algorithm. The algorithm enhances exploitation process of ABC and avoids the premature convergence problem by combining genetic operators into the ABC. These operators are hybridized in two phases of ABC: (1) exploitation process of onlooker bees phase, to enhance information sharing in between employed and onlooker bees which can find better solutions, and (2) scout bees phase to improve the replacement of any exhausted solution. The proposed network model for DNR was compared to some ether established models using different bio-inspired algorithms. The results reveal that the proposed model outperforms the other models in terms of predictive ability and computational time. [ABSTRACT FROM AUTHOR]

Most image inpainting algorithms have problems such as fuzzy images, texture distortion and semantic inaccuracy, and the image inpainting effect is limited when processing photos with large missing sections and resolution levels. The paper proposes an effective image inpainting algorithm via partial multi-scale channel attention mechanism and deep neural networks to address the above phenomenon that existing image inpainting methods using deep learning modules have insufficient perception and representation capabilities for multi-scale features with high proportion of irregular defects. Initially, we used the Res-U-Net module as a generator. The U-Net-like backbone network topology can achieve the encoding and decoding stages of damaged images. Secondly, the residual network structure was built in the encoder and decoder to improve the ability of the proposed network to extract and display the features of the damaged images. Finally, the partial multi-scale channel attention module was inserted in the skip connection with the decoder to increase the efficiency of using the low-level features of the original images. The experimental results of the research can show that the proposed method outperforms state-of-the-art methods in terms of subjective visual perception and objective evaluation indicators on the CelebA , Places2 and Paris Street View datasets. • The model has improved the ability to extract and express multi-scale features. • The method had included partial multi-scale channel attention and residual network. • The backbone network structure can realize encoding and decoding stage. • The structure of residual networks had constructed to enhance the abilities. [ABSTRACT FROM AUTHOR]

The objective of this paper is to simulate the algorithm to detect object for Automated Guide Vehicle (AGV) guidance problem, namely to obstacle avoidance. Currently, the AGV is a transport vehicle widely used in manufacturing factories and plays an important role in the design of material handling system, moving goods to raw materials or finished product to rightful destination that work automatically. This method has been designed for use in security of internal transportation system in order to prevent a collision to the AGVs and the obstacles, which do not know the exact shape, size and color. Such real-time obstacle detection was crucial, we need to classify the obstacles that are real obstacles or fake obstacles. The computer vision with artificial neural networks (ANN) can help the AGV can see and recognize like a human by imitating the functions of the human brain. In this paper, we proposed an algorithm for detecting obstacle that work automatically with no human is required. The algorithm is divided in to two parts: First, a preliminary feature extraction stage by using the Histograms of Oriented Gradients (HOG) to extract height objects feature from the actual video images. Second, the recognition and classification of the obstacles that are real obstacles or fake obstacles such as a painting or text on the floor by a feed-forward time delay neural network (TDNN) in sequences of black and white video images taken by an on-board camera, then decide whether to hold or continue to be moved. As a result, this paper presents the simulate the algorithm to detect the obstacles for AGV, which can work to destination are defined and avoided obstacles by computer vision with TDNN as an alternative principle with high accuracy and reliability. [ABSTRACT FROM AUTHOR]

An accurate battery remaining useful life (RUL) estimation can facilitate the design of a reliable battery system as well as the safety and reliability of actual operation. A reasonable definition and an effective prediction algorithm are indispensable for the achievement of an accurate RUL estimation result. In this paper, the analysis of battery terminal voltage curves under different cycle numbers during charge process is utilized for RUL definition. Moreover, the relationship between RUL and charge curve is simulated by feed forward neural network (FFNN) for its simplicity and effectiveness. Considering the nonlinearity of lithium-ion charge curve, importance sampling (IS) is employed for FFNN input selection. Based on these results, an online approach using FFNN and IS is presented to estimate lithium-ion battery RUL in this paper. Experiments and numerical comparisons are conducted to validate the proposed method. The results show that the FFNN with IS is an accurate estimation method for actual operation. [ABSTRACT FROM AUTHOR]

The paper describes an algorithm for the synthesis of neural networks to control gyrostabilizer. The neural network performs the role of a state vector observer. The role of such an observer is to provide feedback on gyrostabilizer, which is illustrated in the article. The paper details the issue of specific stage-related peculiarities of classic algorithms: choosing the network architecture, learning the neural network and verifying the results of feedback control. The article presents an optimal configuration of the neural network like memory depth, number of layers and neurons in these layers, and activation layer functions. It also provides data on dynamic systems to improve learning neural network learning. There is also provided an optimal training pattern. [ABSTRACT FROM AUTHOR]

This paper introduces a novel PSO-GA based hybrid training algorithm with Adam Optimization and contrasts performance with the generic Gradient Descent based Backpropagation algorithm with Adam Optimization for training Artificial Neural Networks. We aim to overcome the shortcomings of the traditional algorithm, such as slower convergence rate and frequent convergence to local minima, by employing the characteristics of evolutionary algorithms. PSO has a property of faster convergence rate, which can be exploited to account for the slower pace of convergence of the traditional BP (which is due to low values of gradients). In contrast, the integration with GA complements the drawback of convergence to local minima as GA, possesses the capability of efficient global search. So by this integration of these algorithms, we propose our new hybrid algorithm for training ANNs. We compare both the algorithms for the application of medical diagnosis. Results display that the proposed hybrid training algorithm, significantly outperforms the traditional training algorithm, by enhancing the accuracies of the ANNs with an increase of 20% in the average testing accuracy and 0.7% increase in the best testing accuracy. [ABSTRACT FROM AUTHOR]

Anomaly detection is considered as a challenging task due to its imbalanced and unlabelled nature. To overcome this challenge, the combination of different machine learning approaches such as supervised, unsupervised, semi-supervised learning are proposed in the literature. With the advent of neural networks and generative models, different methodologies derived from neural networks are applied to anomaly detection tasks. In this study, we use the KDDCUP99 data set, consider it as an anomaly detection task, and implement BiGAN, considering it as a one-class anomaly detection algorithm. Since generator and discriminator are highly dependent on each other in the training phase, to reduce this dependency, in this paper, we propose two different training approaches for BiGAN by adding extra training steps to it. We also demonstrate that proposed approaches increased the performance of BiGAN on anomaly detection task. [ABSTRACT FROM AUTHOR]

Abstract: In this paper, the AdaBoost-BP algorithm is used to construct a new model to predict the critical frequency of the ionospheric F2-layer (foF2) one hour ahead. Different indices were used to characterize ionospheric diurnal and seasonal variations and their dependence on solar and geomagnetic activity. These indices, together with the current observed foF2 value, were input into the prediction model and the foF2 value at one hour ahead was output. We analyzed twenty-two years’ foF2 data from nine ionosonde stations in the East-Asian sector in this work. The first eleven years’ data were used as a training dataset and the second eleven years’ data were used as a testing dataset. The results show that the performance of AdaBoost-BP is better than those of BP Neural Network (BPNN), Support Vector Regression (SVR) and the IRI model. For example, the AdaBoost-BP prediction absolute error of foF2 at Irkutsk station (a middle latitude station) is 0.32MHz, which is better than 0.34MHz from BPNN, 0.35MHz from SVR and also significantly outperforms the IRI model whose absolute error is 0.64MHz. Meanwhile, AdaBoost-BP prediction absolute error at Taipei station from the low latitude is 0.78MHz, which is better than 0.81MHz from BPNN, 0.81MHz from SVR and 1.37MHz from the IRI model. Finally, the variety characteristics of the AdaBoost-BP prediction error along with seasonal variation, solar activity and latitude variation were also discussed in the paper. [Copyright &y& Elsevier]

Abstract: This paper shows the determination of typical load profiles (TLPs) by using Fuzzy C-Means (FCM) clustering and Probability Neural Network (PNN) classification method. Load profiles provide useful information on electricity consumption to both consumers and suppliers. Precise knowledge on consumer''s electricity consumption becomes essential for the suppliers to design tariffs, load balancing, load planning, etc. As for consumers, such knowledge is important for them to keep track of their electricity consumption and to enable them to take part in retail market especially for those who do not have appropriate metering equipment. Hence, an effective, fast and cheap method to obtain load profile has to be developed. Objectives of this paper are to obtain groups of TLPs by using FCM clustering and to assign TLPs using PNN algorithm. One of the main issues in FCM clustering is the determination of optimal number of cluster. This paper proposes optimal number of cluster determination by using Davies-Bouldin index. PNN network is trained to directly allocate the TLPs to the representative groups of consumers. The data used in this project are obtained from Tenaga Nasional Berhad (TNB). [Copyright &y& Elsevier]

Abstract: The load forecast level in power system is a important symbol to measure operations and management of power system. This paper summarized the research conditions of the short-term load forecasting using artificial neural network method, and analyzed the characteristics of electrical load and factors of influencing power load forecasting accuracy. The paper used the particle swarm optimization neural network method in short-term load forecasting of power grid. Based on the analysis history loads in California power system, we established the load forecasting model considering the various affecting factors, and normalized the input load, meanwhile quantified date, atmosphere and other factors. The example showed that the model of neural network based on the particle swarm optimization algorithm can improve the prediction precision and speed, it''s performance prediction is superior to the model based on BP neural network load forecasting. [Copyright &y& Elsevier]

Background and Objective: The condition of vessel of the human eye is an important factor for the diagnosis of ophthalmological diseases. Vessel segmentation in fundus images is a challenging task due to complex vessel structure, the presence of similar structures such as microaneurysms and hemorrhages, micro-vessel with only one to several pixels wide, and requirements for finer results.Methods: In this paper, we present a multi-scale deeply supervised network with short connections (BTS-DSN) for vessel segmentation. We used short connections to transfer semantic information between side-output layers. Bottom-top short connections pass low level semantic information to high level for refining results in high-level side-outputs, and top-bottom short connection passes much structural information to low level for reducing noises in low-level side-outputs. In addition, we employ cross-training to show that our model is suitable for real world fundus images.Results: The proposed BTS-DSN has been verified on DRIVE, STARE and CHASE_DB1 datasets, and showed competitive performance over other state-of-the-art methods. Specially, with patch level input, the network achieved 0.7891/0.8212 sensitivity, 0.9804/0.9843 specificity, 0.9806/0.9859 AUC, and 0.8249/0.8421 F1-score on DRIVE and STARE, respectively. Moreover, our model behaves better than other methods in cross-training experiments.Conclusions: BTS-DSN achieves competitive performance in vessel segmentation task on three public datasets. It is suitable for vessel segmentation. The source code of our method is available at: https://github.com/guomugong/BTS-DSN. [ABSTRACT FROM AUTHOR]

Highlights • Color image is converted to be grayscale by CNN via gradient domain modeling. • Auxiliary variable principle enables the input and output dimensions to be same. • Final result is reconstructed by global grayscale image and local gradient images. • The relation between color-to-gray and multi-exposure fusion is revealed. • Evaluations on decolorization and multi-exposure fusion are conducted. Abstract This paper shows how to convert a color image to grayscale using convolutional neural network (CNN), that preserves visual contrast via gradient domain modeling. We propose to explore the auxiliary variable principle to make the input and output variable dimensions to be the same, and use L1-norm error of the image gradients as the loss function criterion. The similarity measure calculates the summation of the gradient correlation between each channel of the color image and the transformed grayscale image. The final gray mapping result is then obtained by reconstruction from a globally initial grayscale image and locally derived gradient images. A weighted objective is proposed to balance the robustness and visual appearance of color images. Furthermore, by revealing the relation between color-to-gray and multi-exposure fusion, the network is applied to multi-exposure fusion. Both quantitative and qualitative evaluations on decolorization and multi-exposure fusion consistently demonstrate the potential of the proposed method against existing state-of-the-art algorithms. Graphical abstract Image, graphical abstract [ABSTRACT FROM AUTHOR]

Abstract Nowadays, the research on network mapping is mostly limited to static resource allocation. In fact, the user's demand of network resources changes dynamically over time. Therefore, how to predict the time-varying demand of users and allocate appropriate resources becomes an important way to improve resource utilization. As a fully connected artificial neural network (ANN), the RBFN (Radial Basis Function Network) has diagnostic, predictive and classification functions. However, due to the excessive use of hidden RBF units during training process, it suffers from expensive core inner product calculations and long training time. This paper proposes a dynamic network resource demand predicting algorithm based on RBF incremental design (GSO-INC-RBFDM). In the network mapping, the group search optimizer (GSO) is used to optimize the mapping scheme, and then the radial basis function (RBF) of the incremental construction is used to predict the time-varying demand of the user. GSO-INC-RBFDM based on incremental design of RBF can construct a compact neural network structure, which not only accelerates the training speed, but also improves the predictive accuracy. Simulation experiments show, compared with traditional algorithms and the original RBF, GSO-INC-RBFDM have lower cost, higher acceptance rate and network revenue. [ABSTRACT FROM AUTHOR]

Abstract This paper presents a machine learning-based approach to power transformer fault diagnosis based on dissolved gas analysis (DGA), a bat algorithm (BA), optimizing the probabilistic neural network (PNN). PNN is a radial basis function feedforward neural network based on Bayesian decision theory, which has a strong fault tolerance and significant advantages in pattern classification. However, one challenge still remains: the performance of PNN is greatly affected by its hidden layer element smooth factor which impacts the classification performance. The proposed approach addresses this challenge by deploying the BA algorithm, a kind of bio-inspired algorithm to optimize PNN. Using the real data collected from a transformer system, we conducted the experiments for validating the performance of the developed method. The experimental results demonstrated that BA is an effective algorithm for optimizing PNN smooth factor and BA-PNN can improve the fault diagnosis performance; in turn, and the machine learning-based model (BA-PNN) can significantly enhance the accuracies of power transformer fault diagnosis. [ABSTRACT FROM AUTHOR]

Abstract: This paper, the third one in a three-paper sequence, presents the result of TOSSIM simulation of a Hopfield neural network as a static optimizer and configured to solve the maximum independent set (MIS) problem using a wireless sensor network as a fully parallel and distributed computing hardware platform. TinyOS with its default protocol stack along with nesC were used to develop the simulation model. Simulations were realized for mote counts of 10, 50, 100, and 182; messaging complexity, memory and simulation time costs were measured. Results indicated, as the most prominent finding, that the neural optimization algorithm was able to compute solutions to the MIS problem. The memory footprint of the TOSSIM process in Windows XP environment was about 20 MB for the range of sensor networks considered. The messaging complexity as measured by the total number of messages transmitted and the simulation time increased rather quickly indicating a need to optimize and tune certain aspects of the simulation environment if wireless sensor networks with higher mote counts need to be simulated. [Copyright &y& Elsevier]

Abstract: The financial health of construction contractors is critical in successfully completing a project, and thus default prediction is highly concerned by owners and other stakeholders. In other industries many previous studies employ support vector machine (SVM) or other Artificial Neural Networks (ANN) methods for corporate default prediction using the sample-matching method, which produces sample selection biases. In order to avoid the sample selection biases, this paper used all available firm-years samples during the sample period. Yet this brings a new challenge: the number of non-defaulted samples greatly exceeds the defaulted samples, which is referred to as between-class imbalance. Although the SVM algorithm is a powerful learning process, it cannot always be applied to data with extreme distribution characteristics. This paper proposes an enforced support vector machine-based model (ESVM model) for the default prediction in the construction industry, using all available firm-years data in our sample period to solve the between-class imbalance. The traditional logistic regression model is provided as a benchmark to evaluate the forecasting ability of the ESVM model. All financial variables related to the prediction of contractor default risk as well as 7 variables selected by the Multivariate Discriminant Analysis (MDA) stepwise method are put in the models for comparison. The empirical results of this paper show that the ESVM model always outperforms the logistic regression model, and is more convenient to use because it is relatively independent of the selection of variables. Thus, we recommend the proposed ESVM model as an alternative to the traditionally used logistic model. [Copyright &y& Elsevier]

Abstract: The fish swarm algorithm (FSA) is a new intelligent swarm modeling approach that consists primarily of searching, swarming, and following behaviors. This paper proposes several improvements of the FSA, including: (1) using particle swarm optimization formulation to reformulate the FSA, (2) integrating communication behavior into FSA, and (3) creating formulas for major FSA parameters. This paper also focuses on studying the effects of FSA behaviors on optimization during the evolution process. Results focus on the two case study categories of function optimization (eight benchmark functions) and neural network learning (single-input single-output system identification, multi-inputs single output system identification and Iris classification problem). Evidence indicates that the proposed FSA approach reduces the effort necessary to set parameters and that the proposed communication behavior indeed improves FSA. [Copyright &y& Elsevier]

Abstract: Forest fire occurrence prediction plays a major role in resource allocation, mitigation and recovery efforts. This paper compares two artificial intelligence based methods, artificial neural networks (ANN) and support vector machines (SVM), utilizing a reduced set of weather parameters. Using a reduced set of parameters results in an efficient and reduced cost prediction system especially for developing countries. In this paper the aim is to predict forest fire occurrence by reducing the number of monitored features, and eliminating the need for weather prediction mechanisms. The reason is to reduce errors due to inaccuracies in weather prediction. The challenge is to choose a limited number of easily measurable features in the aim of reducing the cost of the system and its maintenance. At the same time, the chosen features must have a high correlation with the risk of fire occurrence. A literature review of forest fire prediction methods divided into systems/indices, and artificial intelligence is provided. The two fire danger prediction algorithms utilize relative humidity and cumulative precipitation to output a risk estimate. The assessment of these algorithms, using data from Lebanon, demonstrated their ability to accurately predict the risk of fire occurrence on a scale of four levels. [Copyright &y& Elsevier]

Abstract: In this paper, we consider some different aspects involved in the prediction of biological time series. This is often a difficult task, which can be accomplished by defining a suited function approximation problem whose inputs are determined using past samples of the sequence to be predicted. We demonstrate that neural networks can be very useful in this regard. For instance, we illustrate how Mixture of Gaussian neural networks can be efficiently used for the prediction of glucose temporal evolution in Diabetes Mellitus Type 2, instead of using the well-known MINMOD model. We also focus on the chaotic nature of biological time series, which can be contaminated by spurious noise. Neural networks represent a powerful tool for the prediction of these sequences, when the past samples to be used for prediction are chosen by suitable embedding techniques. A new embedding approach is proposed in the paper, using a genetic algorithm where each individual represents a possible embedding solution. We demonstrate that the proposed technique yields better prediction accuracies, considering as a particular application the prediction of those biological parameters (glucose, heart rate, and skin conductivity) that are strictly related to road safety, since they may evidence stress conditions or possible loss of consciousness. [Copyright &y& Elsevier]

Abstract: Process monitoring using multivariate statistical process control (MSPC) has attracted large industries types due to its practical importance and application. In this paper, a combined use of principal component analysis (PCA) and self organisation map (SOM) algorithms are considered. Habitually PCA method uses T2 Hoteling''s and squared predicted error (SPE) as indexes to classify processes variability. In this paper, new version of indexes called metric distances obtained from the self organisation map (SOM) algorithm replace the conventional indexes proper to PCA. A comparative study between SOM, the conventional PCA and the hybrid form of PCA–SOM is examined. Application is made on the real data obtained from a pickling process. As shown in different figures, the combined approach remains important comparatively to PCA but not more than SOM. [Copyright &y& Elsevier]

Abstract: This paper describes a novel binary classification method named LASCUS that can be applied to uneven datasets and sensitive problems such as malfunction detection. Such method aims at filling the gap left by traditional algorithms which have difficulties when coping with unbalanced datasets and are not able to satisfactorily recognize unfrequent patterns. The proposed method is based on the use of a self organizing map (SOM) and of a fuzzy inference system (FIS). The SOM creates a set of clusters to be associated either to frequent or unfrequent situations while the FIS determines such association on the basis of data distribution. The method has been tested on the widely used benchmarking Wisconsin breast cancer database and on two industrial applications. The obtained results, which are discussed in the paper, are encouraging and in line with expectations. [ABSTRACT FROM AUTHOR]

Abstract: This paper describes an approach based on artificial neural network (ANN) to identify the material flow curves of strain hardened 5083-H111 and 5754-O Al–Mg alloys at the temperature ranges from room temperature (RT) to 300°C and a strain rate of 0.0016–0.16s−1. The tensile tests were performed to determine the material responses at various temperatures and strain rates. An ANN model was developed to predict the flow curves of the materials in terms of experimental data. The input parameters of the model are strain rate, temperature, and strain while tensile flow stress is the output. A three layer feed-forward network was trained with BFGS (Broyden, Fletcher, Goldfarb, and Shanno) algorithm. The amount of the neurons in the hidden layer was determined by determining of the root mean square error (RMSE) values for each material. Results reveal that the predicted values in the ANN model are in very good agreement with the experimental data. The ANN model, described in this paper, is an efficient quantitative tool to evaluate and predict the deformation behavior of 5083-H111 and 5754-O Al–Mg alloys for tensile test at prescribed deformation conditions. [ABSTRACT FROM AUTHOR]

Abstract: In this paper, the issue of nonlinear sensitivity analysis for dimensionality reduction in hydrologic model calibration is discussed, and a novel method to quantify the sensitivity of each parameter that considers the nonlinear relationship in the model is presented. The method is based on computing the absolute variation of the nonlinear function represented by the model in its parameter space. The paper discusses the theoretical background of the method and presents the algorithm. The algorithm employs neural network as a pseudo simulator to reduce the computational burden of the analysis. The proposed approach of sensitivity analysis is illustrated through a case study on a physically based distributed hydrologic model. The results indicate that the method is able to rank the parameters effectively, and the ranking can be interpreted in the context of the physical processes being considered by the model. [ABSTRACT FROM AUTHOR]

Abstract: The utilization of motor-in-wheel in the electric vehicle (EV) is a very important research topic. In order to simplify the control of EV, a new system is proposed by Guilin Tao, which changes the traditional configuration and control strategy of direct-wheel-driven EV. In this system two permanent-magnet-brushless dc motors (PMBDCMs) are connected directly in serial, in order to cut down the differential algorithm thoroughly in velocity control strategy. This structure can realize the differential automatically when the EV runs on the smooth road at low speed. But this structure considers the characters of the motor only, which is an ideal system. EV is a complex electromechanical system, and its work condition is also complex. The variation of the external condition and the work condition of operation can affect the operation of motor, so the state of the EV will change. In this paper, according to the change of the work condition and road condition, combined with the dynamics of EV, this problem was well studied. The neural network PID control mode was used in this paper, and by means of the Matlab/Simulink and the experimental verification, it reveals that this method can enhance the safety and stability of this system. [Copyright &y& Elsevier]

Abstract: Based on the universal approximation property of the fuzzy-neural networks, an adaptive fuzzy-neural observer design algorithm is studied for a class of nonlinear SISO systems with both a completely unknown function and an unknown dead-zone input. The fuzzy-neural networks are used to approximate the unknown nonlinear function. Because it is assumed that the system states are unmeasured, an observer needs to be designed to estimate those unmeasured states. In the previous works with the observer design based on the universal approximator, when the dead-zone input appears it is ignored and the stability of the closed-loop system will be affected. In this paper, the proposed algorithm overcomes the affections of dead-zone input for the stability of the systems. Moreover, the dead-zone parameters are assumed to be unknown and will be adjusted adaptively as well as the sign function being introduced to compensate the dead-zone. With the aid of the Lyapunov analysis method, the stability of the closed-loop system is proven. A simulation example is provided to illustrate the feasibility of the control algorithm presented in this paper. [ABSTRACT FROM AUTHOR]

Abstract: The automatic recognition of the modulation format of a detected signal, the intermediate step between signal detection and demodulation, is a major task of an intelligent receiver, with various civilian and military applications. Obviously, with no knowledge of the transmitted data and many unknown parameters at the receiver, such as the signal power, carrier frequency and phase offsets, timing information, etc., blind identification of the modulation is a difficult task. This becomes even more challenging in real world. In this paper I develop a novel algorithm using Two Threshold Sequential Algorithmic Scheme (TTSAS) algorithm and pattern recognition to identify the modulation types of the communication signals automatically. I have proposed and implemented a technique that casts modulation recognition into shape recognition. Constellation diagram is a traditional and powerful tool for design and evaluation of digital modulations. In this paper, modulation classification is performed using constellation of the received signal by fuzzy clustering and consequently hierarchical clustering algorithms are used for classification of Quadrature–Amplitude Modulation (QAM) and Phase Shift Keying (PSK) modulations and also modulated signal symbols constellation utilizing TTSAS clustering algorithm, and matching with standard templates, is used for classification of QAM modulation. TTSAS algorithm used here is implemented by the Hamming neural network. The simulation results show the capability of this method for modulation classification with high accuracy and appropriate convergence in the presence of noise. [Copyright &y& Elsevier]

In this paper, a novel iris feature extraction technique with intelligent classifier is proposed for high performance iris recognition. We use one dimensional circular profile to represent iris features. The reduced and significant features afterward are extracted by Sobel operator and 1-D wavelet transform. So as to improve the accuracy, this paper combines probabilistic neural network (PNN) and particle swarm optimization (PSO) for an optimized PNN classifier model. A comparative experiment of existing methods for iris recognition is evaluated on CASIA iris image databases. The experimental results reveal the proposed algorithm provides superior performance in iris recognition. [Copyright &y& Elsevier]

Abstract: The paper deals with the problem of predicting the time to default in credit behavioural scoring. This area opens a possibility of including a dynamic component in behavioural scoring modelling which enables making decisions related to limit, collection and recovery strategies, retention and attrition, as well as providing an insight into the profitability, pricing or term structure of the loan. In this paper, we compare survival analysis and neural networks in terms of modelling and results. The neural network architecture is designed such that its output is comparable to the survival analysis output. Six neural network models were created, one for each period of default. A radial basis neural network algorithm was used to test all six models. The survival model used a Cox modelling procedure. Further, different performance measures of all models were discussed since even in highly accurate scoring models, misclassification patterns appear. A systematic comparison ‘3+2+2’ procedure is suggested to find the most effective model for a bank. Additionally, the survival analysis model is compared to neural network models according to the relative importance of different variables in predicting the time to default. Although different models can have very similar performance measures they may consist of different variables. The dataset used for the research was collected from a Croatian bank and credit customers were observed during a 12-month period. The paper emphasizes the importance of conducting a detailed comparison procedure while selecting the best model that satisfies the users’ interest. [Copyright &y& Elsevier]

Abstract: This paper outlines an algorithm for incrementally growing Artificial Neural Networks. The algorithm allows the network to expand by adding new sub-networks or modules to an existing structure; the modules are trained using an Evolutionary Algorithm. Only the latest module added to the network is trained, the previous structure remains fixed. The algorithm allows information from different data domains to be integrated into the network and because the search space in each iteration is small, large and complex networks with a modular structure can emerge naturally. The paper describes an application of the algorithm to a legged robot and discusses its biological inspiration. [Copyright &y& Elsevier]

Abstract: This paper presents a new technique of high accuracy to recognize both typewritten and handwritten English and Arabic texts without thinning. After segmenting the text into lines (horizontal segmentation) and the lines into words, it separates the word into its letters. Separating a text line (row) into words and a word into letters is performed by using the region growing technique (implicit segmentation) on the basis of three essential lines in a text row. This saves time as there is no need to skeletonize or to physically isolate letters from the tested word whilst the input data involves only the basic information—the scanned text. The baseline is detected, the word contour is defined and the word is implicitly segmented into its letters according to a novel algorithm described in the paper. The extracted letter with its dots is used as one unit in the system of recognition. It is resized into a 9×9 matrix following bilinear interpolation after applying a lowpass filter to reduce aliasing. Then the elements are scaled to the interval [0,1]. The resulting array is considered as the input to the designed neural network. For typewritten texts, three types of Arabic letter fonts are used—Arial, Arabic Transparent and Simplified Arabic. The results showed an average recognition success rate of 93% for Arabic typewriting. This segmentation approach has also found its application in handwritten text where words are classified with a relatively high recognition rate for both Arabic and English languages. The experiments were performed in MATLAB and have shown promising results that can be a good base for further analysis and considerations of Arabic and other cursive language text recognition as well as English handwritten texts. For English handwritten classification, a success rate of about 80% in average was achieved while for Arabic handwritten text, the algorithm performance was successful in about 90%. The recent results have shown increasing success for both Arabic and English texts. [Copyright &y& Elsevier]

Abstract: The deformation behavior of type 304L stainless steel during hot torsion is investigated using artificial neural network (ANN). Torsion tests in the temperature range of 600–1200°C and in the (maximum surface) strain rate range of 0.1–100s−1 were carried out. These experiments provided the required data for training the neural network and for subsequent testing. The input parameters of the model are strain, log strain rate and temperature while torsional flow stress is the output. A three layer feed-forward network was trained with standard back propagation (BP) and Resilient propagation (Rprop) algorithm. The paper makes a robust comparison of the performances of the above two algorithms. The network trained with Rprop algorithm is found to perform better and also needs less number of iterations for convergence. The developed ANN model employing this algorithm could efficiently track the work hardening, dynamic softening and flow localization regions of the deforming material. Sensitivity analysis showed that temperature and strain rate are the most significant parameters while strain affects the flow stress only moderately. The ANN model, described in this paper, is an efficient quantitative tool to evaluate and predict the deformation behavior of type 304L stainless steel during hot torsion. [Copyright &y& Elsevier]