Artificial Neural Network In Order Research Articles

Privacy Management Solution in Ubiquitous Environments Using Percontrol

The identification of users in pervasive and ubiquitous environments can be performed using various devices and techniques, among which the use of wireless sensor networks (WSN), radio frequency identifiers (RFID), Smartphones, Bluetooth, among others. However, it is necessary to implement management techniques and control privacy of these heterogeneous environments. Thus, we performed a study using artificial neural networks in order to perform the processing of information identified only generating an output of information to Percontrol.

Applied linear discriminant analysis and artificial neural network for sorting dried figs based on texture properties

Dried figs are one of the horticultural products that require sorting in the postharvest stage in order to be presented to the market. In Iran, figs are graded manually by professional workers or automatically by mechanical machines. This paper presents a new algorithm based on machine vision technology applicable to be installed in the fig sorting machines. In the presented methodology, image texture properties of figs are extracted by an image processing algorithm. Some features selected by stepwise linear discriminant analysis were introduced as the superior ones for discriminating different classes of dried figs. Among the ten features, discriminant analysis selected six. The selected texture features were fed to artificial neural networks in order to implement the classification process. The image processing assisted neural networks methodology showed promising result as the total sorting accuracy was 100%.

Improved Short-Term Load Forecasting Based on Two-Stage Predictions with Artificial Neural Networks in a Microgrid Environment

Short-Term Load Forecasting plays a significant role in energy generation planning, and is specially gaining momentum in the emerging Smart Grids environment, which usually presents highly disaggregated scenarios where detailed real-time information is available thanks to Communications and Information Technologies, as it happens for example in the case of microgrids. This paper presents a two stage prediction model based on an Artificial Neural Network in order to allow Short-Term Load Forecasting of the following day in microgrid environment, which first estimates peak and valley values of the demand curve of the day to be forecasted. Those, together with other variables, will make the second stage, forecast of the entire demand curve, more precise than a direct, single-stage forecast. The whole architecture of the model will be presented and the results compared with recent work on the same set of data, and on the same location, obtaining a Mean Absolute Percentage Error of 1.62% against the original 2.47% of the single stage model.

MOIRAE: A computational strategy to extract and represent structural information from experimental protein templates

The prediction and analysis of the three- dimensional (3D) structure of proteins is a key research problem in Structural Bioinformatics. The 1990's Genome Projects resulted in a large increase in the number of available protein sequences. However, the number of identified 3D protein structures have not followed the same growth trend. Currently, the number of available protein sequences greatly exceeds the number of known 3D structures. Many computational methodologies, systems and algorithms have been proposed to address the protein structure prediction problem. However, the problem still remains challenging because of the complexity and high dimensionality of a protein conformational search space. The most significant progress in the last Critical Assessment of protein Structure Prediction was achieved by methods that use database information. Nevertheless, a major challenge remains in the development of better strategies for template identification and representation. This article describes a computational strategy to acquire and represent structural information of experimentally determined 3D protein structures. A clustering strategy was combined with artificial neural networks in order to extract structural information from experimental protein structure templates. In the proposed strategy, the main efforts focus on the acquisition of useful and accurate structural information from 3D protein templates stored in the Protein Data Bank (PDB). The proposed method was tested in twenty protein sequences whose sizes vary from 14 to 70 amino acid residues. Our results show that the proposed method is a good way to extract and represent valuable information obtained from the PDB and also significantly reduce the 3D protein conformational search space.

ニューラルネットワークを用いたリアルタイム高潮予測における学習パラメータの感度分析に関する研究

In the paper, the sensitivity of real-time storm surge forecast models to the input parameters is described based on the artificial neural network in order to predict the storm surge at Sakai Minato in the Sanin coast. In addition, the forecast time spans of 01, 02, 03, 04, 05, 12 and 24 hours are investigated. The study reveals that the real-time forecast model for the 24 forecast successfully predicted the observation at Sakai Minato. The performance of the real time forecast model for the 24h forecast was best when using the data set consisting of the storm surge, the sea level pressure, the depression rate of the sea level pressure and the typhoon position.

Familial or Sporadic Idiopathic Scoliosis - classification based on artificial neural network and GAPDH and ACTB transcription profile.

BackgroundImportance of hereditary factors in the etiology of Idiopathic Scoliosis is widely accepted. In clinical practice some of the IS patients present with positive familial history of the deformity and some do not. Traditionally about 90% of patients have been considered as sporadic cases without familial recurrence. However the exact proportion of Familial and Sporadic Idiopathic Scoliosis is still unknown. Housekeeping genes encode proteins that are usually essential for the maintenance of basic cellular functions. ACTB and GAPDH are two housekeeping genes encoding respectively a cytoskeletal protein β-actin, and glyceraldehyde-3-phosphate dehydrogenase, an enzyme of glycolysis. Although their expression levels can fluctuate between different tissues and persons, human housekeeping genes seem to exhibit a preserved tissue-wide expression ranking order. It was hypothesized that expression ranking order of two representative housekeeping genes ACTB and GAPDH might be disturbed in the tissues of patients with Familial Idiopathic Scoliosis (with positive family history of idiopathic scoliosis) opposed to the patients with no family members affected (Sporadic Idiopathic Scoliosis). An artificial neural network (ANN) was developed that could serve to differentiate between familial and sporadic cases of idiopathic scoliosis based on the expression levels of ACTB and GAPDH in different tissues of scoliotic patients. The aim of the study was to investigate whether the expression levels of ACTB and GAPDH in different tissues of idiopathic scoliosis patients could be used as a source of data for specially developed artificial neural network in order to predict the positive family history of index patient.ResultsThe comparison of developed models showed, that the most satisfactory classification accuracy was achieved for ANN model with 18 nodes in the first hidden layer and 16 nodes in the second hidden layer. The classification accuracy for positive Idiopathic Scoliosis anamnesis only with the expression measurements of ACTB and GAPDH with the use of ANN based on 6-18-16-1 architecture was 8 of 9 (88%). Only in one case the prediction was ambiguous.ConclusionsSpecially designed artificial neural network model proved possible association between expression level of ACTB, GAPDH and positive familial history of Idiopathic Scoliosis.

Comparing Support Vector Machines and Artificial Neural Networks in the Recognition of Steering Angle for Driving of Mobile Robots Through Paths in Plantations

The use of mobile robots turns out to be interesting in activities where the action of human specialist is difficult or dangerous. Mobile robots are often used for the exploration in areas of difficult access, such as rescue operations and space missions, to avoid human experts exposition to risky situations. Mobile robots are also used in agriculture for planting tasks as well as for keeping the application of pesticides within minimal amounts to mitigate environmental pollution. In this paper we present the development of a system to control the navigation of an autonomous mobile robot through tracks in plantations. Track images are used to control robot direction by pre-processing them to extract image features. Such features are then submitted to a support vector machine and an artificial neural network in order to find out the most appropriate route. A comparison of the two approaches was performed to ascertain the one presenting the best outcome. The overall goal of the project to which this work is connected is to develop a real time robot control system to be embedded into a hardware platform. In this paper we report the software implementation of a support vector machine and of an artificial neural network, which so far presented respectively around 93% and 90% accuracy in predicting the appropriate route.

Machine learning methods for the automatic evaluation of exercises on sensor-equipped weight training machines

The present paper proposes pattern recognition techniques for the evaluation of exercises performed on weight training machines equipped with a load cell and rotary encoder for the measurement of essential force and weight displacement characteristics during training. The latter parameters can be used for the implementation of intelligent modeling methods like artificial neural networks in order to assess the exercise technique automatically and provide the athlete with appropriate feedback. First results of the developed classifiers indicate good performance values and high classification rates, demonstrating a significant potential of machine learning routines for the autonomous evaluation of performances on weight machines.

Rule extraction from artificial neural networks to discover causes of quality defects in fabric production

In this paper, a novel classification rule extraction algorithm which has been recently proposed by authors is employed to determine the causes of quality defects in a fabric production facility in terms of predetermined parameters like machine type, warp type etc. The proposed rule extraction algorithm works on the trained artificial neural networks in order to discover the hidden information which is available in the form of connection weights in them. The proposed algorithm is mainly based on a swarm intelligence metaheuristic which is known as Touring Ant Colony Optimization (TACO). The algorithm has a hierarchical structure with two levels. In the first level, a multilayer perceptron type neural network is trained and its weights are extracted. After obtaining the weights, in the second level, the TACO-based algorithm is applied to extract classification rules. The main purpose of the present work is to determine and analyze the most effective parameters on the quality defects in fabric production. The parameters and their levels which give the best quality results are tried to be discovered and evaluated by making use of the proposed algorithm. It is also aimed to compare the accuracy of proposed algorithm with several other rule-based algorithms in order to present its competitiveness.

Effect of contact lenses on the protein composition in tear film: a ProteinChip study

The aim of this study was to analyze and compare the effects of rigid gas permeable and soft contact lenses on the protein composition in the tear film of contact lens wearers. Wearers of soft contact lenses (CL_S, n = 13) and rigid gas permeable contact lenses (CL_H, n = 13) were recruited for this study. Thirteen non-contact lens wearers were also included as the control. Tears were collected using Schirmer strips and frozen until use. The tears were eluted and analyzed on ProteinChips SELDI-TOF (surface-enhanced laser desorption and ionization in time of flight mass spectrometry; Bio-Rad, USA) with different chromatographic surfaces (cationic and anionic exchanger and reversed phase surface). The SELDI spectra were analyzed by multivariate statistical analysis and artificial neural networks in order to find a biomarker panel which differentiates best between the groups. In order to identify protein/peptide peaks from SELDI spectra which showed a significant difference between groups, fractionated tear samples were analyzed using MALDI-TOF MS (matrix-assisted laser desorption/ionization time-of-flight mass spectrometry). For validation of biomarkers, we used an antibody microarray approach. Complex patterns of tear proteins and peptides were detected in the control group and in both contact lens groups. The tear protein composition in both wearers of rigid gas permeable (CL_H) and soft contact lenses (CL_S) differed significantly from protein composition in non-contact lens wearers (p < 0.01). The identification of biomarkers revealed an increase of Protein S100 A8 in the group of wearers of soft contact lenses (CL_S) and a decrease of a main tear protein, lysozyme, in both contact lens groups. The identified biomarker cystatin was upregulated in the group of rigid gas permeable lens wearers (CL_H), whereas the protein intensity of secretoglobin was significantly reduced in this group. Using the microarray approach, detected alterations could be confirmed. Contact lens wear alters the protein profiles in a complex manner. This study demonstrates that significant changes can be found in wearers of soft contact lenses (CL_S) and rigid gas permeable contact lenses (CL_H). Some biomarker intensities are significantly altered only in the group of rigid gas permeable lens wearers (CL_H).

On characterisation of local stress–strain properties in friction stir welded aluminium AA 5083 sheets using micro-tensile specimen testing and instrumented indentation technique

Stress–strain property variations in welded joints are frequently characterized by determining local stress–strain curves across the joint. A commonly and widely accepted method is testing of micro-tensile specimens where the specimens have to be mechanically extracted form the joint and subsequently tested and analysed similar to that of a conventional tensile tests. A recently developed procedure for the determination of local stress–strain properties is based on instrumented indentation testing where the force–depth curves are analysed via artificial neural networks in order to derive local stress–strain curves. In order to verify this new method on welded joints, this method has been applied to a friction stir weld in 3mm thick aluminium AA 5083 sheet material. It turned out that the local stress–strain curves derived by instrumented indentation tests and neural networks are compatible with that obtained from micro-tensile testing. Further on, the article gives advises on a proper tests specimen preparation as well as on the treatment of possible residual stresses.

Pattern recognition and size determination of internal wood defects based on wavelet neural networks

This paper presented the results of a preliminary study on detecting internal wood defects using an ultrasound technique coupled with wavelet transform and artificial neural networks analysis. At room temperature in the laboratory, the type and size of the wood defects in 275 Elm specimens were detected using a RSM-SY5 ultrasonic instrument. The original signals of the Elm specimens were decomposed using wavelet packets, the energy variation of each node in the fifth layer was calculated, and back-propagation artificial neural networks (BP ANN) were trained and employed for wood defect recognition. The energy variation caused by wood defects mostly depends on the degree of the defect's deterioration (i.e., the more serious the wood defect's deterioration, the larger the energy variation). By comparing the energy variation of all 32 node signals in the fifth layer wavelet packet, the variation of the node (5,0) was the largest and contained the most defect information. The node (5,0) was used as the input in back-propagation artificial neural networks in order to detect the type of defects. The accuracy rate for Elm specimens was at least 90%. The same method was used to test the size and position of the hole-defects in Elm specimens with an accuracy rate of at least 80%.

An approach to the calculation of multilayer magnetic shielding using artificial neural network

The electromagnetic interference has many undesired effects to the office equipments and the performance of technological devices. Therefore, electrical devices should have field protection mechanism against magnetic fields. The shielding mechanism prevents external magnetic field emitted from the device to the vicinity area. This work attempts to apply artificial neural network in order to estimate shielded magnetic field for multilayer shielding application. The multilayer magnetic shielding performances of cylindrical shaped ferromagnetic shields under extremely low frequency are investigated. For this purpose, three different ferromagnetic materials in cylindrical shape are chosen for magnetic shielding. The shielding performances of these materials are measured and a selected set of data is used to train and test an artificial neural network. The proposed neural network model achieves an intelligent decision for the shielded magnetic field level based on distance, unshielded magnetic field value, number of shield layer and skin-depth of the shield.

Prediction of MHC class I binding peptides using probability distribution functions

Binding of peptides to specific Major Histo-compatibility Complex (MHC) molecule is important for understanding immunity and has applications to vaccine discovery and design of immunotherapy. Artificial neural networks (ANN) are widely used by predictions tools to classify the peptides as binders or non-binders (BNB). However, the number of known binders to a specific MHC molecule is limited in many cases, which poses a computational challenge for prediction of BNB and hence, needs improvement in learning of ANN. Here, we describe, the application of probability distribution functions to initialize the weights and biases of the artificial neural network in order to predict HLA-A*0201 binders and non-binders. The 10-fold cross validation has been used to validate the results. It is evident from the results that the A(ROC) for 90% of test cases for Weibull, Uniform and Rayleigh distributions is in the range 0.90-1.0. Further, the standard deviation for AROC was minimum for Weibull distribution, and may be used to train the artificial neural network for HLA-A*0201 MHC Class-I binders and non-binders prediction.

Solar energetic particle flux enhancement as a predictor of geomagnetic activity in a neural network‐based model

Coronal mass ejections (CMEs) are believed to be the principal cause of increased geomagnetic activity. They are regarded as being in context of a series of related solar energetic events, such as X‐ray flares (XRAs) accompanied by solar radio bursts (RSPs) and also by solar energetic particle (SEP) flux. Two types of the RSP events are known to be geoeffective, namely, the RSP of type II, interpreted as the signature of shock initiation in the solar corona, and type IV, representing material moving upward in the corona. The SEP events causing geomagnetic response are known to be produced by CME‐driven shocks. In this paper, we use the method of the artificial neural network in order to quantify the geomagnetic response of particular solar events. The data concerning XRAs and RSPs II and/or IV together with their heliographic positions are taken as the input for the neural network. There is a key question posed in our study: can the successfulness of the neural network prediction scheme based solely on the solar disc observations (XRA and RSP) be improved by additional information concerning the SEP flux? To resolve this problem, we chose the SEP events possessing significant enhancement in the 10‐h window, commencing 12 h after the generation of XRAs. In particular, we consider the flux of high‐energy protons with energies over 10 MeV. We have used a chi‐square test to demonstrate that supplying such extra input data improves the neural network prediction scheme.

Diagnóstico do ceratocone baseado no Orbscan com o auxílio de uma rede neural

To evaluate an artificial neural network in order to correctly identify Orbscan II tests of patients with normal and keratoconus corneas. A retrospective analysis included 98 Orbscan II tests of 59 subjects and an artificial neural network was created and trained based on the Java Neural Network 1.1 software. Seventy-three tests (59 normal tests and 14 keratoconus examinations) were applied to train the neural network and 25 eyes were used to test the method (19 normal eyes and 6 cases of keratoconus corneas). Backpropagation method was performed to train the neural network to 5% error and 0.2 learning rate. The trained neural network presented sensibility and specificity of 83 and 100% respectively. Artificial neural network can accurately help clinicians to classify keratoconus in Orbscan II tests.

Prediction of standard particleboard mechanical properties utilizing an artificial neural network and subsequent comparison with a multivariate regression model

The physical properties (specific gravity, moisture content, thickness swelling and water absorption) and mechanical properties (internal bond strength, bending strength and modulus of elasticity) were determined on 93 Spanish-manufactured standard particleboards of different thicknesses selected randomly at the end of the production process. The testing methods of the corresponding European standards (EN) were used, except in the case of the thickness swelling and absorption tests, for which the Spanish UNE standard was used. The thickness and the values obtained for the physical properties were entered into an artificial neural network in order to predict the mechanical properties of the board. The fit was compared with the usual multivariate regression models. The use of a neural network made it possible to obtain the values of bending strength, modulus of elasticity and internal bond strength of the boards utilizing the known data, not only of thickness, moisture content and specific gravity, but also of thickness swelling and water absorption. The neural network proposed is much better adapted to the observed values than any of the multivariate regression models obtained.

Neural Network Based Model for Predicting Housing Market Performance

The United States real estate market is currently facing its worst hit in two decades due to the slowdown of housing sales. The most affected by this decline are real estate investors and home developers who are currently struggling to break-even financially on their investments. For these investors, it is of utmost importance to evaluate the current status of the market and predict its performance over the short-term in order to make appropriate financial decisions. This paper presents the development of artificial neural network based models to support real estate investors and home developers in this critical task. The paper describes the decision variables, design methodology, and the implementation of these models. The models utilize historical market performance data sets to train the artificial neural networks in order to predict unforeseen future performances. An application example is analyzed to demonstrate the model capabilities in analyzing and predicting the market performance. The model testing and validation showed that the error in prediction is in the range between−2% and +2%.

Integrating radial basis function networks with case-based reasoning for product design

This paper presents a case-based design expert system that automatically determines the design values of a product. We focus on the design problem of a shadow mask which is a core component of monitors in the electronics industry. In case-based reasoning (CBR), it is important to retrieve similar cases and adapt them to meet design specifications exactly. Notably, difficulties in automating the adaptation process have prevented designers from being able to use design expert systems easily and efficiently. In this paper, we present a hybrid approach combining CBR and artificial neural networks in order to solve the problems occurring during the adaptation process. We first constructed a radial basis function network (RBFN) composed of representative cases created by K-means clustering. Then, the representative case most similar to the current problem was adjusted using the network. The rationale behind the proposed approach is discussed, and experimental results acquired from real shadow mask design are presented. Using the design expert system, designers can reduce design time and errors and enhance the total quality of design. Furthermore, the expert system facilitates effective sharing of design knowledge among designers.

Design of an artificial neural network for the estimation of the flashover voltage on insulators

This work attempts to apply an artificial neural network in order to estimate the critical flashover voltage on polluted insulators. The artificial neural network uses as input variables the following characteristics of the insulator: diameter, height, creepage distance, form factor and equivalent salt deposit density, and estimates the critical flashover voltage. The data used to train the network and test its performance is derived from experimental measurements and a mathematical model. Various cases have been studied and their results presented separately. Training and testing sets have been modified for each case.