Abstract: Since dealing with high dimensional data is
computationally complex and sometimes even intractable, recently
several feature reductions methods have been developed to reduce
the dimensionality of the data in order to simplify the calculation
analysis in various applications such as text categorization, signal
processing, image retrieval, gene expressions and etc. Among feature
reduction techniques, feature selection is one the most popular
methods due to the preservation of the original features.
In this paper, we propose a new unsupervised feature selection
method which will remove redundant features from the original
feature space by the use of probability density functions of various
features. To show the effectiveness of the proposed method, popular
feature selection methods have been implemented and compared.
Experimental results on the several datasets derived from UCI
repository database, illustrate the effectiveness of our proposed
methods in comparison with the other compared methods in terms of
both classification accuracy and the number of selected features.
Abstract: Stock portfolio selection is a classic problem in finance,
and it involves deciding how to allocate an institution-s or an individual-s
wealth to a number of stocks, with certain investment objectives
(return and risk). In this paper, we adopt the classical Markowitz
mean-variance model and consider an additional common realistic
constraint, namely, the cardinality constraint. Thus, stock portfolio
optimization becomes a mixed-integer quadratic programming problem
and it is difficult to be solved by exact optimization algorithms.
Chemical Reaction Optimization (CRO), which mimics the molecular
interactions in a chemical reaction process, is a population-based
metaheuristic method. Two different types of CRO, named canonical
CRO and Super Molecule-based CRO (S-CRO), are proposed to solve
the stock portfolio selection problem. We test both canonical CRO
and S-CRO on a benchmark and compare their performance under
two criteria: Markowitz efficient frontier (Pareto frontier) and Sharpe
ratio. Computational experiments suggest that S-CRO is promising
in handling the stock portfolio optimization problem.
Abstract: The development of Artificial Neural Networks
(ANNs) is usually a slow process in which the human expert has to
test several architectures until he finds the one that achieves best
results to solve a certain problem. This work presents a new
technique that uses Genetic Programming (GP) for automatically
generating ANNs. To do this, the GP algorithm had to be changed in
order to work with graph structures, so ANNs can be developed. This
technique also allows the obtaining of simplified networks that solve
the problem with a small group of neurons. In order to measure the
performance of the system and to compare the results with other
ANN development methods by means of Evolutionary Computation
(EC) techniques, several tests were performed with problems based
on some of the most used test databases. The results of those
comparisons show that the system achieves good results comparable
with the already existing techniques and, in most of the cases, they
worked better than those techniques.
Abstract: Arc welding is an important joining process widely used in many industrial applications including production of automobile, ships structures and metal tanks. In welding process, the moving electrode causes highly non-uniform temperature distribution that leads to residual stresses and different deviations, especially buckling distortions in thin plates. In order to control the deviations and increase the quality of welded plates, a fixture can be used as a practical and low cost method with high efficiency. In this study, a coupled thermo-mechanical finite element model is coded in the software ANSYS to simulate the behavior of thin plates located by a 3-2-1 positioning system during the welding process. Computational results are compared with recent similar works to validate the finite element models. The agreement between the result of proposed model and other reported data proves that finite element modeling can accurately predict the behavior of welded thin plates.
Abstract: This paper presents an approach for the design of
fuzzy logic power system stabilizers using genetic algorithms. In the
proposed fuzzy expert system, speed deviation and its derivative
have been selected as fuzzy inputs. In this approach the parameters of
the fuzzy logic controllers have been tuned using genetic algorithm.
Incorporation of GA in the design of fuzzy logic power system
stabilizer will add an intelligent dimension to the stabilizer and
significantly reduces computational time in the design process. It is
shown in this paper that the system dynamic performance can be
improved significantly by incorporating a genetic-based searching
mechanism. To demonstrate the robustness of the genetic based
fuzzy logic power system stabilizer (GFLPSS), simulation studies on
multimachine system subjected to small perturbation and three-phase
fault have been carried out. Simulation results show the superiority
and robustness of GA based power system stabilizer as compare to
conventionally tuned controller to enhance system dynamic
performance over a wide range of operating conditions.
Abstract: The increase on the demand of IT resources diverts
the enterprises to use the cloud as a cheap and scalable solution.
Cloud computing promises achieved by using the virtual machine as a
basic unite of computation. However, the virtual machine pre-defined
settings might be not enough to handle jobs QoS requirements. This
paper addresses the problem of mapping jobs have critical start
deadlines to virtual machines that have predefined specifications.
These virtual machines hosted by physical machines and shared a
fixed amount of bandwidth. This paper proposed an algorithm that
uses the idle virtual machines bandwidth to increase the quote of other
virtual machines nominated as executors to urgent jobs. An algorithm
with empirical study have been given to evaluate the impact of the
proposed model on impatient jobs. The results show the importance
of dynamic bandwidth allocation in virtualized environment and its
affect on throughput metric.
Abstract: Biological sequences from different species are called or-thologs if they evolved from a sequence of a common ancestor species and they have the same biological function. Approximations of Kolmogorov complexity or entropy of biological sequences are already well known to be useful in extracting similarity information between such sequences -in the interest, for example, of ortholog detection. As is well known, the exact Kolmogorov complexity is not algorithmically computable. In prac-tice one can approximate it by computable compression methods. How-ever, such compression methods do not provide a good approximation to Kolmogorov complexity for short sequences. Herein is suggested a new ap-proach to overcome the problem that compression approximations may notwork well on short sequences. This approach is inspired by new, conditional computations of Kolmogorov entropy. A main contribution of the empir-ical work described shows the new set of entropy-based machine learning attributes provides good separation between positive (ortholog) and nega-tive (non-ortholog) data - better than with good, previously known alter-natives (which do not employ some means to handle short sequences well).Also empirically compared are the new entropy based attribute set and a number of other, more standard similarity attributes sets commonly used in genomic analysis. The various similarity attributes are evaluated by cross validation, through boosted decision tree induction C5.0, and by Receiver Operating Characteristic (ROC) analysis. The results point to the conclu-sion: the new, entropy based attribute set by itself is not the one giving the best prediction; however, it is the best attribute set for use in improving the other, standard attribute sets when conjoined with them.
Abstract: Neural processors have shown good results for
detecting a certain character in a given input matrix. In this paper, a
new idead to speed up the operation of neural processors for character
detection is presented. Such processors are designed based on cross
correlation in the frequency domain between the input matrix and the
weights of neural networks. This approach is developed to reduce the
computation steps required by these faster neural networks for the
searching process. The principle of divide and conquer strategy is
applied through image decomposition. Each image is divided into
small in size sub-images and then each one is tested separately by
using a single faster neural processor. Furthermore, faster character
detection is obtained by using parallel processing techniques to test the
resulting sub-images at the same time using the same number of faster
neural networks. In contrast to using only faster neural processors, the
speed up ratio is increased with the size of the input image when using
faster neural processors and image decomposition. Moreover, the
problem of local subimage normalization in the frequency domain is
solved. The effect of image normalization on the speed up ratio of
character detection is discussed. Simulation results show that local
subimage normalization through weight normalization is faster than
subimage normalization in the spatial domain. The overall speed up
ratio of the detection process is increased as the normalization of
weights is done off line.
Abstract: A new Meta heuristic approach called "Randomized gravitational emulation search algorithm (RGES)" for solving vertex covering problems has been designed. This algorithm is found upon introducing randomization concept along with the two of the four primary parameters -velocity- and -gravity- in physics. A new heuristic operator is introduced in the domain of RGES to maintain feasibility specifically for the vertex covering problem to yield best solutions. The performance of this algorithm has been evaluated on a large set of benchmark problems from OR-library. Computational results showed that the randomized gravitational emulation search algorithm - based heuristic is capable of producing high quality solutions. The performance of this heuristic when compared with other existing heuristic algorithms is found to be excellent in terms of solution quality.
Abstract: The H.264/AVC video coding standard contains a number of advanced features. Ones of the new features introduced in this standard is the multiple intramode prediction. Its function exploits directional spatial correlation with adjacent block for intra prediction. With this new features, intra coding of H.264/AVC offers a considerably higher improvement in coding efficiency compared to other compression standard, but computational complexity is increased significantly when brut force rate distortion optimization (RDO) algorithm is used. In this paper, we propose a new fast intra prediction mode decision method for the complexity reduction of H.264 video coding. for luma intra prediction, the proposed method consists of two step: in the first step, we make the RDO for four mode of intra 4x4 block, based the distribution of RDO cost of those modes and the idea that the fort correlation with adjacent mode, we select the best mode of intra 4x4 block. In the second step, we based the fact that the dominating direction of a smaller block is similar to that of bigger block, the candidate modes of 8x8 blocks and 16x16 macroblocks are determined. So, in case of chroma intra prediction, the variance of the chroma pixel values is much smaller than that of luma ones, since our proposed uses only the mode DC. Experimental results show that the new fast intra mode decision algorithm increases the speed of intra coding significantly with negligible loss of PSNR.
Abstract: The residue number system (RNS), due to its
properties, is used in applications in which high performance
computation is needed. The carry free nature, which makes the
arithmetic, carry bounded as well as the paralleling facility is the
reason of its capability of high speed rendering. Since carry is not
propagated between the moduli in this system, the performance is
only restricted by the speed of the operations in each modulus. In this
paper a novel method of number representation by use of redundancy
is suggested in which {rn- 2,rn-1,rn} is the reference moduli set
where r=2k+1 and k =1, 2,3,.. This method achieves fast
computations and conversions and makes the circuits of them much
simpler.
Abstract: The use of human hand as a natural interface for humancomputer interaction (HCI) serves as the motivation for research in hand gesture recognition. Vision-based hand gesture recognition involves visual analysis of hand shape, position and/or movement. In this paper, we use the concept of object-based video abstraction for segmenting the frames into video object planes (VOPs), as used in MPEG-4, with each VOP corresponding to one semantically meaningful hand position. Next, the key VOPs are selected on the basis of the amount of change in hand shape – for a given key frame in the sequence the next key frame is the one in which the hand changes its shape significantly. Thus, an entire video clip is transformed into a small number of representative frames that are sufficient to represent a gesture sequence. Subsequently, we model a particular gesture as a sequence of key frames each bearing information about its duration. These constitute a finite state machine. For recognition, the states of the incoming gesture sequence are matched with the states of all different FSMs contained in the database of gesture vocabulary. The core idea of our proposed representation is that redundant frames of the gesture video sequence bear only the temporal information of a gesture and hence discarded for computational efficiency. Experimental results obtained demonstrate the effectiveness of our proposed scheme for key frame extraction, subsequent gesture summarization and finally gesture recognition.
Abstract: This paper presents the results related to the
interference reduction technique in multistage multiuser detector for
asynchronous DS-CDMA system. To meet the real-time
requirements for asynchronous multiuser detection, a bit streaming,
cascade architecture is used. An asynchronous multiuser detection
involves block-based computations and matrix inversions. The paper
covers iterative-based suboptimal schemes that have been studied to
decrease the computational complexity, eliminate the need for matrix
inversions, decreases the execution time, reduces the memory
requirements and uses joint estimation and detection process that
gives better performance than the independent parameter estimation
method. The stages of the iteration use cascaded and bits processed
in a streaming fashion. The simulation has been carried out for
asynchronous DS-CDMA system by varying one parameter, i.e.,
number of users. The simulation result exhibits that system gives
optimum bit error rate (BER) at 3rd stage for 15-users.
Abstract: The Minimum Weighted Vertex Cover (MWVC) problem is a classic graph optimization NP - complete problem. Given an undirected graph G = (V, E) and weighting function defined on the vertex set, the minimum weighted vertex cover problem is to find a vertex set S V whose total weight is minimum subject to every edge of G has at least one end point in S. In this paper an effective algorithm, called Support Ratio Algorithm (SRA), is designed to find the minimum weighted vertex cover of a graph. Computational experiments are designed and conducted to study the performance of our proposed algorithm. Extensive simulation results show that the SRA can yield better solutions than other existing algorithms found in the literature for solving the minimum vertex cover problem.
Abstract: In this paper, by introducing twice continuously differentiable mappings, we develop an interior path following following method, which enables us to give a constructive proof of the general Brouwer fixed point theorem and thus to solve fixed point problems in a class of non-convex sets. Under suitable conditions, a smooth path can be proven to exist. This can lead to an implementable globally convergent algorithm. Several numerical examples are given to illustrate the results of this paper.
Abstract: As the use of registration packages spreads, the number of the aligned image pairs in image databases (either by manual or automatic methods) increases dramatically. These image pairs can serve as a set of training data. Correspondingly, the images that are to be registered serve as testing data. In this paper, a novel medical image registration method is proposed which is based on the a priori knowledge of the expected joint intensity distribution estimated from pre-aligned training images. The goal of the registration is to find the optimal transformation such that the distance between the observed joint intensity distribution obtained from the testing image pair and the expected joint intensity distribution obtained from the corresponding training image pair is minimized. The distance is measured using the divergence measure based on Tsallis entropy. Experimental results show that, compared with the widely-used Shannon mutual information as well as Tsallis mutual information, the proposed method is computationally more efficient without sacrificing registration accuracy.
Abstract: In this paper we propose a Particle Swarm heuristic
optimized Multi-Antenna (MA) system. Efficient MA systems
detection is performed using a robust stochastic evolutionary
computation algorithm based on movement and intelligence of
swarms. This iterative particle swarm optimized (PSO) detector
significantly reduces the computational complexity of conventional
Maximum Likelihood (ML) detection technique. The simulation
results achieved with this proposed MA-PSO detection algorithm
show near optimal performance when compared with ML-MA
receiver. The performance of proposed detector is convincingly
better for higher order modulation schemes and large number of
antennas where conventional ML detector becomes non-practical.
Abstract: This paper presents a hybrid algorithm for solving a timetabling problem, which is commonly encountered in many universities. The problem combines both teacher assignment and course scheduling problems simultaneously, and is presented as a mathematical programming model. However, this problem becomes intractable and it is unlikely that a proven optimal solution can be obtained by an integer programming approach, especially for large problem instances. A hybrid algorithm that combines an integer programming approach, a greedy heuristic and a modified simulated annealing algorithm collaboratively is proposed to solve the problem. Several randomly generated data sets of sizes comparable to that of an institution in Indonesia are solved using the proposed algorithm. Computational results indicate that the algorithm can overcome difficulties of large problem sizes encountered in previous related works.
Abstract: Like any sentient organism, a smart environment
relies first and foremost on sensory data captured from the real
world. The sensory data come from sensor nodes of different
modalities deployed on different locations forming a Wireless Sensor
Network (WSN). Embedding smart sensors in humans has been a
research challenge due to the limitations imposed by these sensors
from computational capabilities to limited power. In this paper, we
first propose a practical WSN application that will enable blind
people to see what their neighboring partners can see. The challenge
is that the actual mapping between the input images to brain pattern
is too complex and not well understood. We also study the
connectivity problem in 3D/2D wireless sensor networks and propose
distributed efficient algorithms to accomplish the required
connectivity of the system. We provide a new connectivity algorithm
CDCA to connect disconnected parts of a network using cooperative
diversity. Through simulations, we analyze the connectivity gains
and energy savings provided by this novel form of cooperative
diversity in WSNs.
Abstract: In Virtual organization, Knowledge Discovery (KD)
service contains distributed data resources and computing grid nodes.
Computational grid is integrated with data grid to form Knowledge
Grid, which implements Apriori algorithm for mining association
rule on grid network. This paper describes development of parallel
and distributed version of Apriori algorithm on Globus Toolkit using
Message Passing Interface extended with Grid Services (MPICHG2).
The creation of Knowledge Grid on top of data and
computational grid is to support decision making in real time
applications. In this paper, the case study describes design and
implementation of local and global mining of frequent item sets. The
experiments were conducted on different configurations of grid
network and computation time was recorded for each operation. We
analyzed our result with various grid configurations and it shows
speedup of computation time is almost superlinear.