Search

A comprehensive systematic study was carried out in order to identify various deep learning methods developed and used for predicting student academic performance. Predicting academic performance allows for the implementation of various preventive and supportive measures earlier in order to improve academic performance and reduce failure and dropout rates. Although machine learning schemes were once popular, deep learning algorithms are now being investigated to solve difficult predictions of student performance in larger datasets with more data attributes. Deep neural network prediction methods with clear modelling and parameter measurements formulated on publicly available and recognised datasets are the focus of the research. Widely used for academic performance prediction, backpropagation algorithms have been trained and tested with various datasets, especially those related to learning management systems (LMS) and massive open online courses (MOOC). The most widely used prediction method appears to be the standard artificial neural network approach. The long-short-term memory (LSTM) approach has been reported to achieve an accuracy of around 87 percent for temporal student performance data. The number of papers that study and improve this method shows that there is a clear rise in deep learning-based academic performance prediction over the last few years. [ABSTRACT FROM AUTHOR]

This article provides a detailed evaluation of cutting-edge artificial intelligence (AI) approaches and metaheuristic algorithms for optimizing wind turbine location inside wind farms. The growing need for renewable energy sources has fueled an increase in research towards efficient and sustainable wind farm designs. To address this challenge, various AI techniques, including genetic algorithms (GA), particle swarm optimization (PSO), simulated annealing, artificial neural networks (ANNs), convolutional neural networks (CNNs), and reinforcement learning, have been explored in combination with metaheuristic algorithms. The goal is to discover optimal sites for turbine placement based on a variety of parameters such as energy output, cost-effectiveness, environmental impact, and geographical restrictions. The paper examines the advantages and disadvantages of each strategy and highlights current breakthroughs in the area. This assessment adds to continuing efforts to optimize wind farm design and promote the use of clean and sustainable energy sources by offering significant insights into current advances. [ABSTRACT FROM AUTHOR]

Deep learning has gained high popularity in the field of image processing and computer vision applications due to its unique feature extraction property. For this characteristic, deep learning networks used to solve different issues in computer vision applications. In this paper the issue has been raised is classification of logo of formal directors in Iraqi government. The paper proposes a multi-layer convolutional neural network (CNN) to classify and recognize these official logos by train the CNN model on several logos. The experimental show the effectiveness of the proposed method to recognize the logo with high accuracy rate about 99.16%. The proposed multi-layers CNN model proves the effectiveness to classify different logos with various conditions. [ABSTRACT FROM AUTHOR]

Deep residual learning frameworks have achieved great success in image classification. This article presents the use of transfer learning which is applied on mango leaf image dataset for its disease's detection. New methodology and training have been used to facilitate the easy and rapid implementation of the mango leaf disease detection system in practice. Proposed system can be used to identify the mango leaf for whether it is healthy or infected with the diseases like anthracnose or red rust. This paper describes all the steps which are considered during the experimentation and design. These steps include leaf image data collection, its preparation, data assessment by agricultural experts, and selection and training of deep neural network architectures. A deep residual framework, residual neural network (ResNET), was used to perform deep convolutional neural network training. ResNETs are easy to optimize and can achieve better accuracies. The experimental results obtained from "ResNET architectures, such as ResNet18, ResNet34, ResNet50, and ResNet101" show the accuracies from 94% to 98%. ResNET18 architecture selected from above for system design as it gives 98% accuracy for mango leaf disease's detection. System will help farmers to identify leaf diseases in quick and efficient manner and facilitate decision-making in this front. [ABSTRACT FROM AUTHOR]

In communication between planes and satellites, Optical Beamforming Networks (OBFNs), which rely on many small and flat Phased Array Antennas (PAAs), need to be tuned in order to receive signals from specific angles. In this paper, we develop a deep neural network representation of tuning OBFNs. The problem of tuning an OBFN is in many aspects similar to training a deep neural network. We present a way to exploit the special structure of OBFNs into deep neural network and an algorithm for tuning OBFNs based on feedback that can be easily measured in real system. Training data, which consists of full signals, can be measured, and therefore is used in this paper. For pilot signals, the desired signal is known explicitly. Given the configuration of OBFNs and all nominal parameters required, it was verified in simulation that the deep neural network can be used to tune large scale OBFNs for any desired delays. [ABSTRACT FROM AUTHOR]

Individuals with hearing and speaking impairment communicate using sign language. The movement of hand, body and expressions of face are the means by which the people, who are unable to hear and speak, can communicate. Bangla sign alphabets are formed with one or two hand movements. There are some features which differentiates the signs. To detect and recognize the signs, analyzing its shape and comparing its features is necessary. This paper aims to propose a model and build a computer systemthat can recognize Bangla Sign Lanugage alphabets and translate them to corresponding Bangla letters by means of deep convolutional neural network (CNN). CNN has been introduced in this model in form of a pre-trained model called "MobileNet" which produced an average accuracy of 95.71% in recognizing 36 Bangla Sign Language alphabets. [ABSTRACT FROM AUTHOR]

The dairy cattle productivity is very depending on the quality of their facilities and their environment condition. A lot of researches have been done in this area, but it is not developing the knowledge of animal dimensions and behaviours correlated with their facilities design. Complexities of dynamics zoometry is very depending on cow behaviours that it forced to use neural network (NN) approaching. Hence, the purposes of the paper is to create the concept of static and dynamic zoometry approach to guide the ergonomics facilities design and the modern of dairy cattle house based on zoometry data by using artificial neural network. The research is started with study literature of anthropometry, CAD design, dairy cattle, facility design and neural network. The next step is collecting the data of cow dimensions for static zoometry (17 variables) and cow behaviours for developing the concept of dynamics zoometry (5 variables). The static data is used as input factor and dynamic data as desire of Back Propagation NN model. The result of BPNN training is used to design the dairy cattle facilities, e.g. 3D CAD house with minimal length = 357.67 cm, minimal width = 132.03 cm (per tail) and minimal height = 205.28 cm. Finally, the paper is successfully developed the concept of zoometry approach and BPNN model as pioneer of implementing comfort knowledge. [ABSTRACT FROM AUTHOR]

Numeral recognition is an important research direction in field of pattern recognition, and it has broad application prospects. Aiming at four arithmetic operations of general printed formats, this article adopts a multiple hybrid recognition method and is applied to automatically calculating. This method mainly uses BP neural network and template matching method to distinguish the numerals and operators, in order to increase the operation speed and recognition accuracy. Sample images of four arithmetic operations are extracted from printed books, and they are used for testing the performance of proposed recognition method. The experiments show that the method provides correct recognition rate of 96% and correct calculation rate of 89%. [ABSTRACT FROM AUTHOR]

The paper demonstrates an enhancement in the mobile robot's performance during trajectory tracking with static obstacles. An adaptive artificial neural network (ANN) control methodology with online tuning evolutionary slice genetic algorithm is used for the motion control of the nonlinear dynamics mobile robot system. This paper aims at locating the optimal path from the starting point to the target point and designing an ANN trajectory tracking control methodology. The algorithm is simulated with fixed-global environment obstacles to demonstate the effectiveness of the ANN controller and the evolutionary optimization algorithm in terms of the shortest path length generated and the minimum number of the evaluation cost function calculated. The simulation results illustrate that the ANN controller's parameters are obtained quickly, generating smooth wheels' torque actions for the mobile robot platform with a minimum cost function evolution that lead to minimize the tracking error to approximately zero with no oscillation in the responses. [ABSTRACT FROM AUTHOR]

The paper tried to integrate the DGA data with the gas production rate, which are the major indexes of transformer fault diagnosis. Duval's triangle method, BP neural network and IEC three-ratio method were weighted. Firstly, the paper regarded the gas production rate as the independent variables, fitted the cubic curves of the gas production rate and variance of each diagnosis method, and then defined the weights of each algorithm through the data processing method of unequal precision. At last, the dynamic weighted combination diagnosis model was established. That is, the weight is different as the gas production rate changes although the method is identical. The results of diagnosis examples show that the accuracy rate of the weighted combination model is higher than any single algorithm, and it has certain stability as well. [ABSTRACT FROM AUTHOR]

Physiological maturity is an important indicator for beef quality. In traditional method, the maturity grade is determined by subjectively evaluating the degree of cartilage ossification at the tips of the dorsal spine of the thoracic vertebrae. This paper uses the computer vision to replace the artificial method for extracting object (cartilage and bone) regions. Hu invariant moments of object region were calculated as the regional shape characteristic parameters. A trained Hopfield neural network model was used for recognizing cartilage and bone area in thoracic vertebrae image based on minimum Euclidean distance. The result showed that the accuracy of network recognition for cartilage and bone region was 92.75% and 87.68%, respectively. For automatically maturity prediction, the accuracy of prediction was 86%. Algorithm proposed in this paper proved the image description and neural network modeling was an effective method for extracting image feature regions. [ABSTRACT FROM AUTHOR]

Image is often subject to noise pollution during the process of collection, acquisition and transmission, noise is a major factor affecting the image quality, which has greatly impeded people from extracting information from the image. The purpose of image denoising is to restore the original image without noise from the noise image, and at the same time maintain the detailed information of the image as much as possible. This paper, by combining artificial bee colony algorithm and BP neural network, proposes the image denoising method based on artificial bee colony and BP neural network (ABC-BPNN), ABC-BPNN adopts the "double circulation" structure during the training process, after specifying the expected convergence speed and precision, it can adjust the rules according to the structure, automatically adjusts the number of neurons, while the weight of the neurons and relevant parameters are determined through bee colony optimization. The simulation result shows that the algorithm proposed in this paper can maintain the image edges and other important features while removing noise, so as to obtain better denoising effect. [ABSTRACT FROM AUTHOR]

Image compression is to compress the redundancy between the pixels as much as possible by using the correlation between the neighborhood pixels so as to reduce the transmission bandwidth and the storage space. This paper applies the integration of wavelet analysis and artificial neural network in the image compression, discusses its performance in the image compression theoretically, analyzes the multi- resolution analysis thought, constructs a wavelet neural network model which is used in the improved image compression and gives the corresponding algorithm. Only the weight in the output layer of the wavelet neural network needs training while the weight of the input layer can be determined according to the relationship between the interval of the sampling points and the interval of the compactly-supported intervals. Once determined, training is unnecessary, in this way, it accelerates the training speed of the wavelet neural network and solves the problem that it is difficult to determine the nodes of the hidden layer in the traditional neural network. The computer simulation experiment shows that the algorithm of this paper has more excellent compression effect than the traditional neural network method. [ABSTRACT FROM AUTHOR]

In order to solve being sensitive to the initial weights, slow convergence, being easy to fall into local minimum and other problems of the BP neural network, this paper introduces the Particle Swarm Optimization algorithm into the Artificial Neural Network training, and construct a BP neural network model optimized by the particle swarm optimization. This method can speed up the convergence and improve the prediction accuracy. Through the analysis of the main factors on the cost of transmission line project, dig out the path and lead factors, topography and meteorological factors, the tower and the tower base materials and other factors. Use the PSO-BP model for the cost forecasting of transmission line project based on historical project data. The result shows that the method can predict the cost effectively. Compared with the traditional BP neural network, the method can predict with higher accuracy, and can be generalized and applied in cost forecasting of actual projects. [ABSTRACT FROM AUTHOR]

This paper describes a design of adaptive liquid level control system using the concept of Multi Sensor Data Fusion (MSDF). Purpose of the work is to design a controller for accurately controlling the level of liquid in a process tank with liquid temperature changes. The proposed objective is obtained by i) implementing a MSDF framework using Pau's framework for measuring liquid level and temperature, ii) analyzing the behavior of actuator output for variation in liquid temperature, and iii) designing a suitable adaptive controller which will produce desired control action for controlling liquid level accurately using neural network algorithms. Outputs from sensors are fused to obtain the fluid level output and also relation of level transmitter output for change in temperature. This information is used by controller to train the neural network so as to tune the controller parameters (proportional gain, integral constant, and differential constant), to drive the actuator. Results obtained show that the system is able to control liquid level within range of 1.915% of set point even with variations in liquid temperature. [ABSTRACT FROM AUTHOR]

Liquid flow and level control are essential requirements in various industries, such as paper manufacturing, petrochemical industries, waste management, and others. Controlling the liquids flow and levels in such industries is challenging due to the existence of nonlinearity and modeling uncertainties of the plants. This paper presents a method to control the liquid level in a second tank of a coupled-tank plant through variable manipulation of a water pump in the first tank. The optimum controller parameters of this plant are calculated using radial basis function neural network metamodel. A time-varying nonlinear dynamic model is developed and the corresponding linearized perturbation models are derived from the nonlinear model. The performance of the developed optimized controller using metamodeling is compared with the original large space design. In addition, linearized perturbation models are derived from the nonlinear dynamic model with time-varying parameters. [ABSTRACT FROM AUTHOR]

Because of increasing requirements of the present consumers and industrial units utilizing sensitive loads, there is need of good power quality in order to retain the power quality standards. Nowadays the study of the voltage flicker is becoming essential part of power quality studies. The flicker is typically the effect of a rapidly changing load which is large with respect to the short circuit ability of an electrical supply system. The inferior effects of voltage flicker include malfunctioning of power electronic equipment. Also it causes annoying effects to human. Hence detection of the flicker source is an essential step in the power quality assessment process. This paper delivers a review about methods used to identify flicker producing loads in accordance with IEC 61000-4-15. Once the report related to the disturbance place is known, an investigation and corrective action can be accordingly carried out. Also a method based upon Discrete Wavelet Transform and Artificial Neural Network is proposed to detect initial instance of occurrence of flicker. [ABSTRACT FROM AUTHOR]

An adaptive radical basis function (RBF) neural network PID control scheme for X-Y position table is proposed by the paper. Firstly, X-Y position table model is established, controller based on neutral network is used to learn adaptive and compensate uncertainty model of X-Y position table, neutral network is used to study model. PID neural network controller base on augmented variable method is designed. PID controller is used as assistant direction error controller, neural network parameters base on stochastic gradient algorithm can be adjust adaptive on line. The simulation results show that the presented controller has important engineering value. [ABSTRACT FROM AUTHOR]

In recent years, many machine learning methods have been used in network traffic identification. In order to improve the accuracy and solve some problems of network traffic identification, this paper presents a multi layer perceptron neural network-based method for network traffic identification, and parameters of multi-layer perceptron neural network are analyzed. Experimental results show that this method can effectively solve some problems, and can improve the classification correctness. [ABSTRACT FROM AUTHOR]

Automatic identity recognition of ear images represents an active area of interest within the biometric community. The human ear is a perfect source of data for passive person identification. Ear images can be captured from a distance and in a covert manner; this makes ear recognition technology an attractive choice for security applications and surveillance in addition to related application domains. Differing from other biometric modalities, the human ear is neither affected by expressions like faces are nor do need closer touching like fingerprints do. In this paper, a deep learning object detector called faster region based convolutional neural networks (Faster R-CNN) is used for ear detection. A convolutional neural network (CNN) is used as feature extraction. principal component analysis (PCA) and genetic algorithm are used for feature reduction and selection respectively and a fully connected artificial neural network as a matcher. The testing proved the accuracy of 97.8% percentage of success with acceptable speed and it confirmed the accuracy and robustness of the proposed system. [ABSTRACT FROM AUTHOR]

This paper develops an adaptive control coordination scheme for power system stabilizers (PSSs) to improve the oscillation damping and dynamic performance of interconnected multimachine power system. The scheme was based on the use of a neural network which identifies online the optimal controller parameters. The inputs to the neural network include the active and reactive- power of the synchronous generators which represent the power loading on the system and elements of the reduced nodal impedance matrix for representing the power system configuration. The outputs of the neural network were the parameters of the PSSs which lead to optimal oscillation damping for the prevailing system configuration and operating condition. For a representative power system, the neural network has been trained and tested for a wide range of credible operating conditions and contingencies. Both eigenvalue calculations and time-domain simulations were used in the testing and verification of the performance of the neural network-based stabilizer. [ABSTRACT FROM AUTHOR]

The settlement monitoring of subway runs through the entire construction stage of subway. It is very important to predict the accurate settlement value for construction safety of subway. In this paper, the wavelet transform is used to denoise the settlement data. The auxiliary wavelet neural network, embedded wavelet neural network and single BP neural network are applied to predict the settlement of Tianjin subway. Compared with single BP neural network and auxiliary wavelet neural network, the embedded wavelet neural network model has a higher accuracy and better prediction effect. The embedded wavelet neural network is more valuable than the BP neural network model so it can be used in the prediction of subway settlement. [ABSTRACT FROM AUTHOR]

Copyright of Telkomnika is the property of Department of Electrical Engineering, Ahmad Dahlan University and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Telkomnika is the property of Department of Electrical Engineering, Ahmad Dahlan University and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Telkomnika is the property of Department of Electrical Engineering, Ahmad Dahlan University and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

This paper proposes a face recognition module built as a web service. We introduce a novel design and mechanism for face recognition on a web platform and to memorize most recent users for the user. This web service is called Vision and developed using the Flask and TensorFlow deep learning framework. The face recognition process is powered by FaceNet deep convolutional network model. The face recognition process done by Vision could also be utilized for user authentication and user memorization, both done in on a web platform. As a demonstration of concept and viability, in this study, Vision is integrated into a web-based voice chatbot. The testing and evaluation of Vision's face recognition process show an overall F-score of one for all test scenarios. [ABSTRACT FROM AUTHOR]

Automated medical image processing, particularly of radiological images, can reduce the number of diagnostic errors, increase patient care and reduce medical costs. This paper seeks to evaluate the performance of three recent convolutional neural networks in the autonomous identification of fissures over two-dimensional radiological images. These architectures have been proposed as deep neural network types specially designed for image classification, which allows their integration with traditional image processing strategies for automatic analysis of medical images. In particular, we use three convolutional networks: ResNet (residual neural network), DenseNet (dense convolutional network), and NASNet (neural architecture search network) to learn information from a set of 200 images labeled half as fissured bones and half as seamless bones. All three networks are trained and adjusted under the same conditions, and their performance was evaluated with the same metrics. The final results consider not only the model's ability to predict the characteristics of an unknown image but also its internal complexity. The three neural models were optimized to reduce classification errors without producing network over-adjustment. In all three cases, generalization of behavior was observed, and the ability of the models to identify the images with fissures, however the expected performance was only achieved with the NASNet model. [ABSTRACT FROM AUTHOR]

Human face recognition (HFR) is the method of recognizing people in images or videos. There are different HFR methods such as feature-based, eigen-faces, hidden markov model and neural network (NN) based methods. Feature extraction or preprocessing used in first three mentioned methods that associated with the category of the image to recognize. While in the NN method, any type of image can be useful without the requirement to particular data about the type of image, and simultaneously provides superior accuracy. In this paper, HFR system based on neural-fuzzy (NF) has been introduced. In the NN system, backpropagation (BP) algorithm is used to update the weights of the neurons through supervised learning. Two sets of the image have been used for training and testing the network to identify the person. If the test image matches to one of the trained sets of the image, then the system will return recognized. And if the test image does not match to one of the trained sets of the image, then the system will return not recognized. The feature extraction methods used in this paper is Geometric moments and Color feature extraction. The recognition rate of 95.556 % has been achieved. The experimental result illustrations that the association of two techniques that provide better accuracy. [ABSTRACT FROM AUTHOR]

This paper presents a simple technique to perform inverse halftoning using the deep learning framework. The proposed method inherits the usability and superiority of deep residual learning to reconstruct the halftone image into the continuous-tone representation. It involves a series of convolution operations and activation function in forms of residual block elements. We investigate the usage of pre-activation function and standard activation function in each residual block. The experimental section validates the proposed method ability to effectively reconstruct the halftone image. This section also exhibits the proposed method superiority in the inverse halftoning task compared to that of the handcrafted feature schemes and former deep learning approaches. The proposed method achieves 30.37 dB and 0.9481 on the average peak signal-to-noise ratio (PSNR) and structural similarity index (SSIM) scores, respectively. It gives the improvements around 1.67 dB and 0.0481 for those values compared to the most competing scheme. [ABSTRACT FROM AUTHOR]

Strabismus is one of the widespread vision disorders in which the eyes are misaligned and asymmetric. Convolutional neural networks (CNNs) are properly designed for analyzing images and detecting texture patterns. In this paper, we proposed a system that uses deep learning CNN applications for automatically detecting and classifying strabismus disorder. The proposed system includes two main stages: first, the detection of facial eye segmentation using the viola-jones algorithm. The second stage is to map the segmented eye area according to the iris position of each eye. This method is applied to three strabismus datasets, gathered as digital images. The second section covers the segmentation of the eye region. Besides, the evaluation equations for measuring system performance. The system has undergone numerous experiments in various stages to simulate and analyze the detection performance of CNN layers through different classifiers and variant thresholds ratio. The researchers investigated the experimental outcomes during the training and testing phases and obtained promising results that exhibit the effectiveness of the proposed system. According to the results, the accuracy of this technique reached 95.62%. [ABSTRACT FROM AUTHOR]

The traditional RBF neural network has the problem of slow training speed and low efficiency, this paper puts forward the algorithm of improvement of RBF neural network by AdaBoost algorithm combined with PSO, to expand the application range of the RBF neural network. Firstly, it preprocesses the sample data in training set, and initialize the weights of test data; Secondly, it optimizes and chooses different implied layer functions and network learning parameters by using the improved PSO algorithm, to produce different types of RBF weak predictor, and repeatedly train the sample data by using Matlab tools; Finally, it constructs multiple generated RBF weak predictors to strong predictors by using AdaBoost iterative algorithm. It chooses data sets of UCI database to do the simulation experiment, and the simulation results show that the proposed algorithm further reduces the mean absolute error, compared with the traditional RBF neural network prediction, the experiment has improved the prediction precision of the network, to provide a reference for RBF neural network prediction. [ABSTRACT FROM AUTHOR]

In this paper, the drawbacks of current harmonic frequency estimation algorithm in power system are pointed out. In order to improve the detection accuracy and reduce the computational complexity, a new harmonic estimation algorithm for power system is proposed by combining the ESPRIT algorithm with linear neural network. Theoretic analysis and simulation experiments indicated that the new algorithm has super resolution characteristics, ideally suited for harmonic analysis in power system. [ABSTRACT FROM AUTHOR]

This paper introduces the basic principle and calculation steps of BP neural network algorithm for recognition of odor characteristics. Using the PEN3 electronic nose collects the volatile components of milk. Simulation of BP neural network algorithm in MATLAB condition, and compared with the partial least squares identification algorithm which is in the software of electronic nose. It is verified that the odor characteristic parameters with nonlinear and little differences could be better identified by BP neural network. [ABSTRACT FROM AUTHOR]

Neural network is one of the branches with the most active research, development and application in computational intelligence and machine study. Radial basis function neural network (RBFNN) has achieved some success in more than one application field, especially in pattern recognition and functional approximation. Due to its simple structure, fast training speed and excellent generalization ability, it has been widely used. Artificial fish swarm algorithm (AFSA) is a new swarm intelligent optimization algorithm derived from the study on the preying behavior of fish swarm. This algorithm is not sensitive to the initial value and the parameter selection, but strong in robustness and simple and easy to realize and it also has parallel processing capability and global searching ability. This paper mainly researches the weight and threshold of AFSA in optimizing RBFNN. The simulation experiment proves that AFSA-RBFNN is significantly advantageous in global optimization capability and that it has outstanding global optimization ability and stability. [ABSTRACT FROM AUTHOR]

In the current era, researchers have been active in confirming and achieving their work through simulation using the computer program Matlab, in addition to the comparison between different control methods is also one of the prevailing behaviors, and the focus has been on the use of electrical machines in industry through multiple applications. Researchers in this study selected type of electric motors and two types of control systems for comparison, and to verify the possibility of improving the system's work performance through the simulation results, the process of achieving the objectives of the current research is carried out. This paper presents using conventional proportional-integral-derivative (PID) controller and artificial neural networks (ANN) with direct current servo motor (DCSM) in order to obtain good performance characteristics because of efficient and widely use of this motor in the fields of control. The motor model in addition to the controller is built using Matlab simulation software. A comparison was made between these controllers (PID and ANN), where the simulation results indicate that the neural networks being developmental in the process of simulating the operation of the servo motor type and got good performance and better results from the traditional real-time console use case. [ABSTRACT FROM AUTHOR]

Size fitting problem is a main obstacle to large scale online garment sales. It is the difficult to customers to find the fit garments when they couldn't try on. In this paper, we present an algorithm base on wavelet neural network to help customer choosing their clothing specifications automatically. After the reasonable wavelet function is selected, we established the model structure and the initial parameters. The wavelet neural network is trained by the body measures and the result of AHP algorithm after normalization. The new data are used to test the network. As a result, the error from wavelet neural network is smaller, and the prediction accuracy is proved than that from the algorithm based on traditional BP network. [ABSTRACT FROM AUTHOR]

Currently, the method of optimizing the wavelet neural network with particle swarm plays a certain role in improving the convergence speed and accuracy; however, it is not a good solution for problems of turning into local extrema and poor global search ability. To solve these problems, this paper, based on the particle swarm optimization, puts forward an improved method, which is introducing the chaos mechanism into the algorithm of chaotic particle swarm optimization. Through a series of comparative simulation experiments, it proves that applying this algorithm to optimize the wavelet neural network can successfully solve the problems of turning into local extrema, and improve the convergence speed of the network, in the meantime, reduce the output error and improve the search ability of the algorithm. In general, it helps a lot to improve the overall performance of the wavelet neural network. [ABSTRACT FROM AUTHOR]

A new complex-valued wavelet neural network is proposed in this paper, by introducing a modified complex-valued back propagation algorithm, in which a new error function is to be minimized by the algorithm. The improvement performance is further confirmed by the simulation results, which show that the modified algorithm is simpler than the conventional algorithm, and has better convergence, better stability and faster running speed. [ABSTRACT FROM AUTHOR]

This paper proposes the combined methods of Wavelet Transform (WT) and Euclidean Distance (ED) to estimate the expected value of the possibly feature vector of Indonesian syllables. This research aims to find the best properties in effectiveness and efficiency on performing feature extraction of each syllable sound to be applied in the speech recognition systems. This proposed approach which is the state-of-the-art of the previous study consist of three main phase. In the first phase, the speech signal is segmented and normalized. In the second phase, the signal is transformed into frequency domain by using the WT. In the third phase, to estimate the expected feature vector, the ED algorithm is used. Th e result shows the list of features of each syllables can be used for the next research, and some recommendations on the most effective and efficient WT to be used in performing syllable sound recognition. [ABSTRACT FROM AUTHOR]

Substrate Integrated Waveguide (SIW) antennas are considered as main radiators for RF and microwave wireless systems due to their low profile, low cost and soft integration with the other devices. The gain of a SIW patch antenna may be enhanced using different techniques such as Artificial Neural Network s (ANN) b y modifying the antenna’s geometry with high efficiency comparing to electromagnetic techniques that take more time. This paper describes a novel structure of a circular SIW patch antenna design using a tree-dimensional electromagnetic (3D-EM) simulation based on ANN model which is developed as an accurate tool for synthesizing the forward side and then analyzing the reverse side of the problem. In this work, ANN algorithms are used for training the samples to provide precise geometrical dimensions of the SIW patch antenna with high accuracy for the target requirements. The antenna is designed to operate in Ku and K frequency bands, resonate at 16.10 GHz and 19.81 GHz respectively and show good performance resulting in low return losses of less than -10dB to -29dB for the selective frequency bands. [ABSTRACT FROM AUTHOR]

Polluted insulators at high voltages has acquired considerable importance with the rise of voltage transmission lines. The contamination may lead to flashover voltage. As a result, flashover voltage could lead to service outage and affects negatively the reliability of the power system. This paper presents a dynamic model of ac 50Hz flashover voltages of polluted hydrophobic polymer insulators. The models are constructed using the regression tree method, artificial neural network (ANN), and adaptive neuro -fuzzy (ANFIS). For this purpose, more than 2000 different experimental testing conditions were used to generate a training set. The study of the ac flashover voltages depends on silicone rubber (SiR) percentage content in ethylene propylene diene monomer (EPDM) rubber. Besides, water conductivity (μS/cm), number of droplets on the surface, and volume of water droplet (ml) are considered. The regression tree model is obtained and the performance of the proposed system with other intelligence methods is compared. It can be concluded that the performance of the least squares regression tree model outperforms the other intelligence methods, which gives the proposed model better generalization ability. [ABSTRACT FROM AUTHOR]

Copyright of Telkomnika is the property of Department of Electrical Engineering, Ahmad Dahlan University and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Telkomnika is the property of Department of Electrical Engineering, Ahmad Dahlan University and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Telkomnika is the property of Department of Electrical Engineering, Ahmad Dahlan University and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Telkomnika is the property of Department of Electrical Engineering, Ahmad Dahlan University and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

This paper discusses the formation of an appropriate regression model in precipitation prediction. Precipitation prediction has a major influence to multiply the agricultural production of potatoes in Tengger, East Java, Indonesia. Periodically, the precipitation has non-linear patterns. By using a non-linear approach, the prediction of precipitation produces more accurate results. Genetic algorithm (GA) functioning chooses precipitation period which forms the best model. To prevent early convergence, testing the best combination value of crossover rate and mutation rate is done. To test the accuracy of the predicted results are used Root Mean Square Error (RMSE) as a benchmark. Based on the RMSE value of each method on every location, prediction using GA-Non-Linear Regression is better than Fuzzy Tsukamoto for each location. Compared to Generalized Space-Time Autoregressive-Seemingly Unrelated Regression (GSTAR-SUR), precipitation prediction using GA is better. This has been proved that for 3 locations GA is superior and on 1 location, GA has the least value of deviation level. [ABSTRACT FROM AUTHOR]

An open circuit (O-C) fault detection method for grid-connected neutral-point-clamped (NPC) inverter based on independent component analysis (ICA) and neural network (NN) is proposed in this paper. A NN classifier is applied to the fault diagnosis of NPC inverter. The ICA is utilized for the three phase current feature extraction. The ICA reduces the number of NN input neuron. A lower dimensional input space reduces the noise and the training time of NN, the ICA algorithm improves the mapping performance. The proposed algorithm is evaluated with simulation test set. The overall classification performance of the proposed network is more than 97%. The simulation results show that the proposed algorithm performs satisfactorily to fault location. [ABSTRACT FROM AUTHOR]

In this paper, we propose a construction of fuzzy radial basis function neural network model for diagnosing prostate cancer. A fuzzy radial basis function neural network (fuzzy RBFNN) is a hybrid model of logical fuzzy and neural network. The fuzzy membership function of the fuzzy RBFNN model input is developed using the triangle function. The fuzzy C-means method is applied to estimate the center and the width parameters of the radial basis function. The weight estimation is performed by various ways to gain the most accurate model. A singular value decomposition (SVD) is exploited to address this process. As a comparison, we perform other ways including back propagation and global ridge regression. The study also promotes image preprocessing using high frequency emphasis filter (HFEF) and histogram equalization (HE) to enhance the quality of the prostate radiograph. The features of the textural image are extracted using the gray level cooccurrence matrix (GLCM) and gray level run length matrix (GLRLM). The experiment results of fuzzy RBFNN are compared to those of RBFNN model. Generally, the performances of fuzzy RBFNN surpass the RBFNN in all accuracy calculation. In addition, the fuzzy RBFNN-SVD demonstrates the most accurate model for prostate cancer diagnosis. [ABSTRACT FROM AUTHOR]

One of the world's leading causes of death is traffic accidents. Data from World Health Organization (WHO) that there are 1.25 million people in the world die each year. Meanwhile, based on data obtained from Statistics Indonesia, traffic accidents from 2006 to 2013 continue to increase. Of all these accidents, the largest accident occurred at motorcyclists, especially motorcyclists who not wearing standard helmet. In controlling the motorcyclists, police view directly at the highway, so that there are weaknesses which there are still a possibility of motorcyclist offenders who are undetectable especially for motorcyclists who are not wear helmet. This paper explains research on image classification of human head wearing a helmet and not wearing a helmet with backpropagation neural network algorithm. The test results of this analysis is the application can performs classification with 86.67% accuracy rate. This research can be developed into a larger system and integrated that can be used to detect motorcyclists who are not wearing helmet. [ABSTRACT FROM AUTHOR]

A novel method for eliminating noise from a noised speech signal in order to improve its quality using combined power spectral subtraction and multi-layer perceptron network is presented in this paper. Firstly, the contaminated speech signal was processed by spectral subtraction to enhance the clean speech signal. Then, the signal was processed by a neural network using the spectral subtraction parameters and result of estimated speech signal in order to improve its signal quality and intelligibility. The artificial neural network used was multi-layer perceptron network consisted of three layers with six input and one output. The neural network was trained with three speech signals contaminated with two level white gaussian noises in SNR including 0 dB and 30dB. The designed speech enhancement was examined with ten noised speech signals. Based on the experiments, the improvement of signal quality SNR was up to 7 dB when the signal quality input was 0dB. Then, based on the PESQ score, the proposed method can improve up to 0.4 from its origin value. Those experiment results show that the proposed method is capable to improve both the signal quality and intelligibility better than the original power spectral subtraction. [ABSTRACT FROM AUTHOR]