Abstract: In this paper, we describe an application for face
recognition. Many studies have used local descriptors to characterize
a face, the performance of these local descriptors remain low by
global descriptors (working on the entire image). The application of
local descriptors (cutting image into blocks) must be able to store
both the advantages of global and local methods in the Discrete
Cosine Transform (DCT) domain. This system uses neural network
techniques. The letter method provides a good compromise between
the two approaches in terms of simplifying of calculation and
classifying performance. Finally, we compare our results with those
obtained from other local and global conventional approaches.
Abstract: In the context of the handwriting recognition, we
propose an off line system for the recognition of the Arabic
handwritten words of the Algerian departments. The study is based
mainly on the evaluation of neural network performances, trained
with the gradient back propagation algorithm. The used parameters to
form the input vector of the neural network are extracted on the
binary images of the handwritten word by several methods. The
Distribution parameters, the centered moments of the different
projections of the different segments, the centered moments of the
word image coding according to the directions of Freeman, and the
Barr features applied binary image of the word and on its different
segments. The classification is achieved by a multi layers perceptron.
A detailed experiment is carried and satisfactory recognition results
are reported.
Abstract: In this paper, we present an optimization technique or
a learning algorithm using the hybrid architecture by combining the
most popular sequence recognition models such as Recurrent Neural
Networks (RNNs) and Hidden Markov models (HMMs). In order to
improve the sequence/pattern recognition/classification performance
by applying a hybrid/neural symbolic approach, a gradient descent
learning algorithm is developed using the Real Time Recurrent
Learning of Recurrent Neural Network for processing the knowledge
represented in trained Hidden Markov Models. The developed hybrid
algorithm is implemented on automata theory as a sample test beds
and the performance of the designed algorithm is demonstrated and
evaluated on learning the deterministic finite state automata.
Abstract: Since the last decade, there has been a rapid growth in
digital multimedia, such as high-resolution media files and threedimentional
movies. Hence, there is a need for large digital storage
such as Hard Disk Drive (HDD). As such, users expect to have a
quieter HDD in their laptop. In this paper, a jury test has been
conducted on a group of 34 people where 17 of them are students
who are the potential consumer, and the remaining are engineers who
know the HDD. A total 13 HDD sound samples have been selected
from over hundred HDD noise recordings. These samples are
selected based on an agreed subjective feeling. The samples are
played to the participants using head acoustic playback system, which
enabled them to experience as similar as possible the same
environment as have been recorded. Analysis has been conducted and
the obtained results have indicated different group has different
perception over the noises. Two neural network-based acoustic
annoyance models are established based on back propagation neural
network. Four psychoacoustic metrics, loudness, sharpness,
roughness and fluctuation strength, are used as the input of the
model, and the subjective evaluation results are taken as the output.
The developed models are reasonably accurate in simulating both
training and test samples.
Abstract: The knitted fabric suffers a deformation in its
dimensions due to stretching and tension factors, transverse and
longitudinal respectively, during the process in rectilinear knitting
machines so it performs a dry relaxation shrinkage procedure and
thermal action of prefixed to obtain stable conditions in the knitting.
This paper presents a dry relaxation shrinkage prediction of Bordeaux
fiber using a feed forward neural network and linear regression
models. Six operational alternatives of shrinkage were predicted. A
comparison of the results was performed finding neural network
models with higher levels of explanation of the variability and
prediction. The presence of different reposes is included. The models
were obtained through a neural toolbox of Matlab and Minitab
software with real data in a knitting company of Southern
Guanajuato. The results allow predicting dry relaxation shrinkage of
each alternative operation.
Abstract: In this paper, we present a comparative study of three
methods of 2D face recognition system such as: Iso-Geodesic Curves
(IGC), Geodesic Distance (GD) and Geodesic-Intensity Histogram
(GIH). These approaches are based on computing of geodesic
distance between points of facial surface and between facial curves.
In this study we represented the image at gray level as a 2D surface in
a 3D space, with the third coordinate proportional to the intensity
values of pixels. In the classifying step, we use: Neural Networks
(NN), K-Nearest Neighbor (KNN) and Support Vector Machines
(SVM). The images used in our experiments are from two wellknown
databases of face images ORL and YaleB. ORL data base was
used to evaluate the performance of methods under conditions where
the pose and sample size are varied, and the database YaleB was used
to examine the performance of the systems when the facial
expressions and lighting are varied.
Abstract: The work reported through this paper is an
experimental work conducted on High Performance Concrete (HPC)
with super plasticizer with the aim to develop some models suitable
for prediction of compressive strength of HPC mixes. In this study,
the effect of varying proportions of fly ash (0% to 50% @ 10%
increment) on compressive strength of high performance concrete has
been evaluated. The mix designs studied were M30, M40 and M50 to
compare the effect of fly ash addition on the properties of these
concrete mixes. In all eighteen concrete mixes that have been
designed, three were conventional concretes for three grades under
discussion and fifteen were HPC with fly ash with varying
percentages of fly ash. The concrete mix designing has been done in
accordance with Indian standard recommended guidelines. All the
concrete mixes have been studied in terms of compressive strength at
7 days, 28 days, 90 days, and 365 days. All the materials used have
been kept same throughout the study to get a perfect comparison of
values of results. The models for compressive strength prediction
have been developed using Linear Regression method (LR), Artificial
Neural Network (ANN) and Leave-One-Out Validation (LOOV)
methods.
Abstract: Predicting earnings management is vital for the capital
market participants, financial analysts and managers. The aim of this
research is attempting to respond to this query: Is there a significant
difference between the regression model and neural networks’
models in predicting earnings management, and which one leads to a
superior prediction of it? In approaching this question, a Linear
Regression (LR) model was compared with two neural networks
including Multi-Layer Perceptron (MLP), and Generalized
Regression Neural Network (GRNN). The population of this study
includes 94 listed companies in Tehran Stock Exchange (TSE)
market from 2003 to 2011. After the results of all models were
acquired, ANOVA was exerted to test the hypotheses. In general, the
summary of statistical results showed that the precision of GRNN did
not exhibit a significant difference in comparison with MLP. In
addition, the mean square error of the MLP and GRNN showed a
significant difference with the multi variable LR model. These
findings support the notion of nonlinear behavior of the earnings
management. Therefore, it is more appropriate for capital market
participants to analyze earnings management based upon neural
networks techniques, and not to adopt linear regression models.
Abstract: Evolution strategy (ES) is a well-known instance of evolutionary algorithms, and there have been many studies on ES. In this paper, the author proposes an extended ES for solving fuzzy-valued optimization problems. In the proposed ES, genotype values are not real numbers but fuzzy numbers. Evolutionary processes in the ES are extended so that it can handle genotype instances with fuzzy numbers. In this study, the proposed method is experimentally applied to the evolution of neural networks with fuzzy weights and biases. Results reveal that fuzzy neural networks evolved using the proposed ES with fuzzy genotype values can model hidden target fuzzy functions even though no training data are explicitly provided. Next, the proposed method is evaluated in terms of variations in specifying fuzzy numbers as genotype values. One of the mostly adopted fuzzy numbers is a symmetric triangular one that can be specified by its lower and upper bounds (LU) or its center and width (CW). Experimental results revealed that the LU model contributed better to the fuzzy ES than the CW model, which indicates that the LU model should be adopted in future applications of the proposed method.
Abstract: This paper suggests a new internal architecture of
holon based on feature selection model using the combination of
Bees Algorithm (BA) and Artificial Neural Network (ANN). BA is
used to generate features while ANN is used as a classifier to
evaluate the produced features. Proposed system is applied on the
Wine dataset, the statistical result proves that the proposed system is
effective and has the ability to choose informative features with high
accuracy.
Abstract: In this study, an Artificial Neural Network (ANN)
analytical method has been developed for analyzing earthquake
performances of the Reinforced Concrete (RC) buildings. 66 RC
buildings with four to ten storeys were subjected to performance
analysis according to the parameters which are the existing material,
loading and geometrical characteristics of the buildings. The selected
parameters have been thought to be effective on the performance of
RC buildings. In the performance analyses stage of the study, level of
performance possible to be shown by these buildings in case of an
earthquake was determined on the basis of the 4-grade performance
levels specified in Turkish Earthquake Code-2007 (TEC-2007). After
obtaining the 4-grade performance level, selected 23 parameters of
each building have been matched with the performance level. In this
stage, ANN-based fast evaluation algorithm mentioned above made
an economic and rapid evaluation of four to ten storey RC buildings.
According to the study, the prediction accuracy of ANN has been
found about 74%.
Abstract: In this paper, we provided a literature survey on the
artificial stock problem (ASM). The paper began by exploring the
complexity of the stock market and the needs for ASM. ASM
aims to investigate the link between individual behaviors (micro
level) and financial market dynamics (macro level). The variety of
patterns at the macro level is a function of the AFM complexity. The
financial market system is a complex system where the relationship
between the micro and macro level cannot be captured analytically.
Computational approaches, such as simulation, are expected to
comprehend this connection. Agent-based simulation is a simulation
technique commonly used to build AFMs. The paper proceeds by
discussing the components of the ASM. We consider the roles
of behavioral finance (BF) alongside the traditionally risk-averse
assumption in the construction of agent’s attributes. Also, the
influence of social networks in the developing of agents interactions is
addressed. Network topologies such as a small world, distance-based,
and scale-free networks may be utilized to outline economic
collaborations. In addition, the primary methods for developing
agents learning and adaptive abilities have been summarized.
These incorporated approach such as Genetic Algorithm, Genetic
Programming, Artificial neural network and Reinforcement Learning.
In addition, the most common statistical properties (the stylized facts)
of stock that are used for calibration and validation of ASM are
discussed. Besides, we have reviewed the major related previous
studies and categorize the utilized approaches as a part of these
studies. Finally, research directions and potential research questions
are argued. The research directions of ASM may focus on the macro
level by analyzing the market dynamic or on the micro level by
investigating the wealth distributions of the agents.
Abstract: Voting algorithms are extensively used to make
decisions in fault tolerant systems where each redundant module
gives inconsistent outputs. Popular voting algorithms include
majority voting, weighted voting, and inexact majority voters. Each
of these techniques suffers from scenarios where agreements do not
exist for the given voter inputs. This has been successfully overcome
in literature using fuzzy theory. Our previous work concentrated on a
neuro-fuzzy algorithm where training using the neuro system
substantially improved the prediction result of the voting system.
Weight training of Neural Network is sub-optimal. This study
proposes to optimize the weights of the Neural Network using
Artificial Bee Colony algorithm. Experimental results show the
proposed system improves the decision making of the voting
algorithms.
Abstract: Human beings have the ability to make logical
decisions. Although human decision - making is often optimal, it is
insufficient when huge amount of data is to be classified. Medical
dataset is a vital ingredient used in predicting patient’s health
condition. In other to have the best prediction, there calls for most
suitable machine learning algorithms. This work compared the
performance of Artificial Neural Network (ANN) and Decision Tree
Algorithms (DTA) as regards to some performance metrics using
diabetes data. WEKA software was used for the implementation of
the algorithms. Multilayer Perceptron (MLP) and Radial Basis
Function (RBF) were the two algorithms used for ANN, while
RegTree and LADTree algorithms were the DTA models used. From
the results obtained, DTA performed better than ANN. The Root
Mean Squared Error (RMSE) of MLP is 0.3913 that of RBF is
0.3625, that of RepTree is 0.3174 and that of LADTree is 0.3206
respectively.
Abstract: DNA Barcode provides good sources of needed
information to classify living species. The classification problem has
to be supported with reliable methods and algorithms. To analyze
species regions or entire genomes, it becomes necessary to use the
similarity sequence methods. A large set of sequences can be
simultaneously compared using Multiple Sequence Alignment which
is known to be NP-complete. However, all the used methods are still
computationally very expensive and require significant computational
infrastructure. Our goal is to build predictive models that are highly
accurate and interpretable. In fact, our method permits to avoid the
complex problem of form and structure in different classes of
organisms. The empirical data and their classification performances
are compared with other methods. Evenly, in this study, we present
our system which is consisted of three phases. The first one, is called
transformation, is composed of three sub steps; Electron-Ion
Interaction Pseudopotential (EIIP) for the codification of DNA
Barcodes, Fourier Transform and Power Spectrum Signal Processing.
Moreover, the second phase step is an approximation; it is
empowered by the use of Multi Library Wavelet Neural Networks
(MLWNN). Finally, the third one, is called the classification of DNA
Barcodes, is realized by applying the algorithm of hierarchical
classification.
Abstract: Artificial neural networks have gained a lot of interest
as empirical models for their powerful representational capacity,
multi input and output mapping characteristics. In fact, most feedforward
networks with nonlinear nodal functions have been proved to
be universal approximates. In this paper, we propose a new
supervised method for color image classification based on selforganizing
feature maps (SOFM). This algorithm is based on
competitive learning. The method partitions the input space using
self-organizing feature maps to introduce the concept of local
neighborhoods. Our image classification system entered into RGB
image. Experiments with simulated data showed that separability of
classes increased when increasing training time. In additional, the
result shows proposed algorithms are effective for color image
classification.
Abstract: This study investigates the effects of the lead angle
and chip thickness variation on surface roughness during the
machining of compacted graphite iron using ceramic cutting tools
under dry cutting conditions. Analytical models were developed for
predicting the surface roughness values of the specimens after the
face milling process. Experimental data was collected and imported
to the artificial neural network model. A multilayer perceptron model
was used with the back propagation algorithm employing the input
parameters of lead angle, cutting speed and feed rate in connection
with chip thickness. Furthermore, analysis of variance was employed
to determine the effects of the cutting parameters on surface
roughness. Artificial neural network and regression analysis were
used to predict surface roughness. The values thus predicted were
compared with the collected experimental data, and the
corresponding percentage error was computed. Analysis results
revealed that the lead angle is the dominant factor affecting surface
roughness. Experimental results indicated an improvement in the
surface roughness value with decreasing lead angle value from 88° to
45°.
Abstract: Several parameters are established in order to measure
biodiesel quality. One of them is the iodine value, which is an
important parameter that measures the total unsaturation within a
mixture of fatty acids. Limitation of unsaturated fatty acids is
necessary since warming of higher quantity of these ones ends in
either formation of deposits inside the motor or damage of lubricant.
Determination of iodine value by official procedure tends to be very
laborious, with high costs and toxicity of the reagents, this study uses
artificial neural network (ANN) in order to predict the iodine value
property as an alternative to these problems. The methodology of
development of networks used 13 esters of fatty acids in the input
with convergence algorithms of back propagation of back
propagation type were optimized in order to get an architecture of
prediction of iodine value. This study allowed us to demonstrate the
neural networks’ ability to learn the correlation between biodiesel
quality properties, in this caseiodine value, and the molecular
structures that make it up. The model developed in the study reached
a correlation coefficient (R) of 0.99 for both network validation and
network simulation, with Levenberg-Maquardt algorithm.
Abstract: This paper presents an application of Artificial Neural
Network (ANN) algorithm for improving power system voltage
stability. The training data is obtained by solving several normal and
abnormal conditions using the Linear Programming technique. The
selected objective function gives minimum deviation of the reactive
power control variables, which leads to the maximization of
minimum Eigen value of load flow Jacobian. The considered reactive
power control variables are switchable VAR compensators, OLTC
transformers and excitation of generators. The method has been
implemented on a modified IEEE 30-bus test system. The results
obtain from the test clearly show that the trained neural network is
capable of improving the voltage stability in power system with a
high level of precision and speed.
Abstract: In this study, three robust predicting methods, namely artificial neural network (ANN), adaptive neuro fuzzy inference system (ANFIS) and support vector machine (SVM) were used for computing the resonant frequency of A-shaped compact microstrip antennas (ACMAs) operating at UHF band. Firstly, the resonant frequencies of 144 ACMAs with various dimensions and electrical parameters were simulated with the help of IE3D™ based on method of moment (MoM). The ANN, ANFIS and SVM models for computing the resonant frequency were then built by considering the simulation data. 124 simulated ACMAs were utilized for training and the remaining 20 ACMAs were used for testing the ANN, ANFIS and SVM models. The performance of the ANN, ANFIS and SVM models are compared in the training and test process. The average percentage errors (APE) regarding the computed resonant frequencies for training of the ANN, ANFIS and SVM were obtained as 0.457%, 0.399% and 0.600%, respectively. The constructed models were then tested and APE values as 0.601% for ANN, 0.744% for ANFIS and 0.623% for SVM were achieved. The results obtained here show that ANN, ANFIS and SVM methods can be successfully applied to compute the resonant frequency of ACMAs, since they are useful and versatile methods that yield accurate results.