Abstract: In this paper we designed and implemented a new
ensemble of classifiers based on a sequence of classifiers which were
specialized in regions of the training dataset where errors of its
trained homologous are concentrated. In order to separate this
regions, and to determine the aptitude of each classifier to properly
respond to a new case, it was used another set of classifiers built
hierarchically. We explored a selection based variant to combine the
base classifiers. We validated this model with different base
classifiers using 37 training datasets. It was carried out a statistical
comparison of these models with the well known Bagging and
Boosting, obtaining significantly superior results with the
hierarchical ensemble using Multilayer Perceptron as base classifier.
Therefore, we demonstrated the efficacy of the proposed ensemble,
as well as its applicability to general problems.
Abstract: In this paper, an improved ant colony optimization
(ACO) algorithm is proposed to enhance the performance of global
optimum search. The strategy of the proposed algorithm has the
capability of fuzzy pheromone updating, adaptive parameter tuning,
and mechanism resetting. The proposed method is utilized to tune the
parameters of the fuzzy controller for a real beam and ball system.
Simulation and experimental results indicate that better performance
can be achieved compared to the conventional ACO algorithms in the
aspect of convergence speed and accuracy.
Abstract: Information is a critical asset and an important source for gaining competitive advantage in firms. The effective maintenance of IT becomes an important task. In order to better understand the determinants of IT effectiveness, this study employs the Industrial Organization (I/O) and Resource Based View (RBV) theories and investigates the industry effect and several major firmspecific factors in relation to their impact on firms- IT effectiveness. The data consist of a panel data of ten-year observations of firms whose IT excellence had been recognized by the CIO Magazine. The non-profit organizations were deliberately excluded, as explained later. The results showed that the effectiveness of IT management varied significantly across industries. Industry also moderated the effects of firm demographic factors such as size and age on IT effectiveness. Surprisingly, R & D investment intensity had negative correlation to IT effectiveness. For managers and practitioners, this study offers some insights for evaluation criteria and expectation for IT project success. Finally, the empirical results indicate that the sustainability of IT effectiveness appears to be short in duration.
Abstract: In this paper we address the problem of musical style
classification, which has a number of applications like indexing in
musical databases or automatic composition systems. Starting from
MIDI files of real-world improvisations, we extract the melody track
and cut it into overlapping segments of equal length. From these
fragments, some numerical features are extracted as descriptors of
style samples. We show that a standard Bayesian classifier can be
conveniently employed to build an effective musical style classifier,
once this set of features has been extracted from musical data.
Preliminary experimental results show the effectiveness of the
developed classifier that represents the first component of a musical
audio retrieval system
Abstract: Brand loyalty is a strategic asset of the company. In
the era of competition to have loyal customers decides on the market
superiority of enterprises. Creating the loyalty of buyers, however, is
a lengthy process and requires the appropriate business strategy,
preceded by the proper market research. The purpose of the paper is
to present the concept of brand loyalty, the creation of loyalty of
customers, the benefits and determinants of loyalty on the example of
brewery market in Poland.
Abstract: Star graphs are Cayley graphs of symmetric groups of permutations, with transpositions as the generating sets. A star graph is a preferred interconnection network topology to a hypercube for its ability to connect a greater number of nodes with lower degree. However, an attractive property of the hypercube is that it has a Hamiltonian decomposition, i.e. its edges can be partitioned into disjoint Hamiltonian cycles, and therefore a simple routing can be found in the case of an edge failure. The existence of Hamiltonian cycles in Cayley graphs has been known for some time. So far, there are no published results on the much stronger condition of the existence of Hamiltonian decompositions. In this paper, we give a construction of a Hamiltonian decomposition of the star graph 5-star of degree 4, by defining an automorphism for 5-star and a Hamiltonian cycle which is edge-disjoint with its image under the automorphism.
Abstract: Recently, information security has become a key issue
in information technology as the number of computer security
breaches are exposed to an increasing number of security threats. A
variety of intrusion detection systems (IDS) have been employed for
protecting computers and networks from malicious network-based or
host-based attacks by using traditional statistical methods to new data
mining approaches in last decades. However, today's commercially
available intrusion detection systems are signature-based that are not
capable of detecting unknown attacks. In this paper, we present a
new learning algorithm for anomaly based network intrusion
detection system using decision tree algorithm that distinguishes
attacks from normal behaviors and identifies different types of
intrusions. Experimental results on the KDD99 benchmark network
intrusion detection dataset demonstrate that the proposed learning
algorithm achieved 98% detection rate (DR) in comparison with
other existing methods.
Abstract: This paper presents an algebraic approach to optimize
queries in domain-specific database management system
for protein structure data. The approach involves the introduction of
several protein structure specific algebraic operators to query the
complex data stored in an object-oriented database system. The
Protein Algebra provides an extensible set of high-level Genomic
Data Types and Protein Data Types along with a comprehensive
collection of appropriate genomic and protein functions. The paper
also presents a query translator that converts high-level query
specifications in algebra into low-level query specifications in
Protein-QL, a query language designed to query protein structure
data. The query transformation process uses a Protein Ontology that
serves the purpose of a dictionary.
Abstract: A fusion classifier composed of two modules, one made by a hidden Markov model (HMM) and the other by a support vector machine (SVM), is proposed to recognize faces with pose variations in open-set recognition settings. The HMM module captures the evolution of facial features across a subject-s face using the subject-s facial images only, without referencing to the faces of others. Because of the captured evolutionary process of facial features, the HMM module retains certain robustness against pose variations, yielding low false rejection rates (FRR) for recognizing faces across poses. This is, however, on the price of poor false acceptance rates (FAR) when recognizing other faces because it is built upon withinclass samples only. The SVM module in the proposed model is developed following a special design able to substantially diminish the FAR and further lower down the FRR. The proposed fusion classifier has been evaluated in performance using the CMU PIE database, and proven effective for open-set face recognition with pose variations. Experiments have also shown that it outperforms the face classifier made by HMM or SVM alone.
Abstract: Bendability is constrained by maximum top roller
load imparting capacity of the machine. Maximum load is
encountered during the edge pre-bending stage of roller bending.
Capacity of 3-roller plate bending machine is specified by
maximum thickness and minimum shell diameter combinations that
can be pre-bend for given plate material of maximum width.
Commercially available plate width or width of the plate that can be
accommodated on machine decides the maximum rolling width.
Original equipment manufacturers (OEM) provide the machine
capacity chart based on reference material considering perfectly
plastic material model. Reported work shows the bendability analysis
of heavy duty 3-roller plate bending machine. The input variables for
the industry are plate thickness, shell diameter and material property
parameters, as it is fixed by the design. Analytical models of
equivalent thickness, equivalent width and maximum width based on
power law material model were derived to study the bendability.
Equation of maximum width provides bendability for designed
configuration i.e. material property, shell diameter and thickness
combinations within the machine limitations. Equivalent thicknesses
based on perfectly plastic and power law material model were
compared for four different materials grades of C-Mn steel in order
to predict the bend-ability. Effect of top roller offset on the
bendability at maximum top roller load imparting capacity is
reported.
Abstract: Problems on algebraical polynomials appear in many fields of mathematics and computer science. Especially the task of determining the roots of polynomials has been frequently investigated.Nonetheless, the task of locating the zeros of complex polynomials is still challenging. In this paper we deal with the location of zeros of univariate complex polynomials. We prove some novel upper bounds for the moduli of the zeros of complex polynomials. That means, we provide disks in the complex plane where all zeros of a complex polynomial are situated. Such bounds are extremely useful for obtaining a priori assertations regarding the location of zeros of polynomials. Based on the proven bounds and a test set of polynomials, we present an experimental study to examine which bound is optimal.
Abstract: Nowadays, with the emerging of the new applications
like robot control in image processing, artificial vision for visual
servoing is a rapidly growing discipline and Human-machine
interaction plays a significant role for controlling the robot. This
paper presents a new algorithm based on spatio-temporal volumes for
visual servoing aims to control robots. In this algorithm, after
applying necessary pre-processing on video frames, a spatio-temporal
volume is constructed for each gesture and feature vector is extracted.
These volumes are then analyzed for matching in two consecutive
stages. For hand gesture recognition and classification we tested
different classifiers including k-Nearest neighbor, learning vector
quantization and back propagation neural networks. We tested the
proposed algorithm with the collected data set and results showed the
correct gesture recognition rate of 99.58 percent. We also tested the
algorithm with noisy images and algorithm showed the correct
recognition rate of 97.92 percent in noisy images.
Abstract: Several works regarding facial recognition have dealt with methods which identify isolated characteristics of the face or with templates which encompass several regions of it. In this paper a new technique which approaches the problem holistically dispensing with the need to identify geometrical characteristics or regions of the face is introduced. The characterization of a face is achieved by randomly sampling selected attributes of the pixels of its image. From this information we construct a set of data, which correspond to the values of low frequencies, gradient, entropy and another several characteristics of pixel of the image. Generating a set of “p" variables. The multivariate data set with different polynomials minimizing the data fitness error in the minimax sense (L∞ - Norm) is approximated. With the use of a Genetic Algorithm (GA) it is able to circumvent the problem of dimensionality inherent to higher degree polynomial approximations. The GA yields the degree and values of a set of coefficients of the polynomials approximating of the image of a face. By finding a family of characteristic polynomials from several variables (pixel characteristics) for each face (say Fi ) in the data base through a resampling process the system in use, is trained. A face (say F ) is recognized by finding its characteristic polynomials and using an AdaBoost Classifier from F -s polynomials to each of the Fi -s polynomials. The winner is the polynomial family closer to F -s corresponding to target face in data base.
Abstract: In this paper, we propose a novel frequency offset
estimation scheme for orthogonal frequency division multiplexing
(OFDM) systems. By correlating the OFDM signals within the coherence
phase bandwidth and employing a threshold in the frequency
offset estimation process, the proposed scheme is not only robust to
the timing offset but also has a reduced complexity compared with
that of the conventional scheme. Moreover, a timing offset estimation
scheme is also proposed as the next stage of the proposed frequency
offset estimation. Numerical results show that the proposed scheme
can estimate frequency offset with lower computational complexity
and does not require additional memory while maintaining the same
level of estimation performance.
Abstract: Faults in a network may take various forms such as hardware/software errors, vertex/edge faults, etc. Folded hypercube is a well-known variation of the hypercube structure and can be constructed from a hypercube by adding a link to every pair of nodes with complementary addresses. Let FFv (respectively, FFe) be the set of faulty nodes (respectively, faulty links) in an n-dimensional folded hypercube FQn. Hsieh et al. have shown that FQn - FFv - FFe for n ≥ 3 contains a fault-free cycle of length at least 2n -2|FFv|, under the constraints that (1) |FFv| + |FFe| ≤ 2n - 4 and (2) every node in FQn is incident to at least two fault-free links. In this paper, we further consider the constraints |FFv| + |FFe| ≤ 2n - 3. We prove that FQn - FFv - FFe for n ≥ 5 still has a fault-free cycle of length at least 2n - 2|FFv|, under the constraints : (1) |FFv| + |FFe| ≤ 2n - 3, (2) |FFe| ≥ n + 2, and (3) every vertex is still incident with at least two links.
Abstract: In the recent works related with mixture discriminant
analysis (MDA), expectation and maximization (EM) algorithm is
used to estimate parameters of Gaussian mixtures. But, initial values
of EM algorithm affect the final parameters- estimates. Also, when
EM algorithm is applied two times, for the same data set, it can be
give different results for the estimate of parameters and this affect the
classification accuracy of MDA. Forthcoming this problem, we use
Self Organizing Mixture Network (SOMN) algorithm to estimate
parameters of Gaussians mixtures in MDA that SOMN is more robust
when random the initial values of the parameters are used [5]. We
show effectiveness of this method on popular simulated waveform
datasets and real glass data set.
Abstract: Clustering algorithms help to understand the hidden
information present in datasets. A dataset may contain intrinsic and
nested clusters, the detection of which is of utmost importance. This
paper presents a Distributed Grid-based Density Clustering algorithm
capable of identifying arbitrary shaped embedded clusters as well as
multi-density clusters over large spatial datasets. For handling
massive datasets, we implemented our method using a 'sharednothing'
architecture where multiple computers are interconnected
over a network. Experimental results are reported to establish the
superiority of the technique in terms of scale-up, speedup as well as
cluster quality.
Abstract: In this paper the concept of the cosets of an anti Lfuzzy
normal subgroup of a group is given. Furthermore, the group
G/A of cosets of an anti L-fuzzy normal subgroup A of a group
G is shown to be isomorphic to a factor group of G in a natural
way. Finally, we prove that if f : G1 -→ G2 is an epimorphism of
groups, then there is a one-to-one order-preserving correspondence
between the anti L-fuzzy normal subgroups of G2 and those of G1
which are constant on the kernel of f.
Abstract: In the present work, behavior of inoxydable steel as
reinforcement bar in composite concrete is being investigated. The
bar-concrete adherence in reinforced concrete (RC) beam is studied
and focus is made on the tension stiffening parameter. This study
highlighted an approach to observe this interaction behavior in
bending test instead of direct tension as per reported in many
references. The approach resembles actual loading condition of the
structural RC beam. The tension stiffening properties are then
applied to numerical finite element analysis (FEA) to verify their
correlation with laboratory results. Comparison with laboratory
shows a good correlation between the two. The experimental settings
is able to determine tension stiffening parameters in RC beam and
the modeling strategies made in ABAQUS can closely represent the
actual condition. Tension stiffening model used can represent the
interaction properties between inoxydable steel and concrete.
Abstract: In this paper, multilayered coreless printed circuit
board (PCB) step-down power transformers for DC-DC converter
applications have been designed, manufactured and evaluated. A set
of two different circular spiral step-down transformers were
fabricated in the four layered PCB. These transformers have been
modelled with the assistance of high frequency equivalent circuit and
characterized with both sinusoidal and square wave excitation. This
paper provides the comparative results of these two different
transformers in terms of their resistances, self, leakage, mutual
inductances, coupling coefficient and also their energy efficiencies.
The operating regions for optimal performance of these transformers
for power transfer applications are determined. These transformers
were tested for the output power levels of about 30 Watts within the
input voltage range of 12-50 Vrms. The energy efficiency for these
step down transformers is observed to be in the range of 90%-97% in
MHz frequency region.