Abstract: Margin-Based Principle has been proposed for a long
time, it has been proved that this principle could reduce the
structural risk and improve the performance in both theoretical
and practical aspects. Meanwhile, feed-forward neural network is
a traditional classifier, which is very hot at present with a deeper
architecture. However, the training algorithm of feed-forward neural
network is developed and generated from Widrow-Hoff Principle that
means to minimize the squared error. In this paper, we propose
a new training algorithm for feed-forward neural networks based
on Margin-Based Principle, which could effectively promote the
accuracy and generalization ability of neural network classifiers
with less labelled samples and flexible network. We have conducted
experiments on four UCI open datasets and achieved good results
as expected. In conclusion, our model could handle more sparse
labelled and more high-dimension dataset in a high accuracy while
modification from old ANN method to our method is easy and almost
free of work.
Abstract: Networking is important among students to achieve
better understanding. Social networking plays an important role in the
education. Realizing its huge potential, various organizations,
including institutions of higher learning have moved to the area of
social networks to interact with their students especially through
Facebook. Therefore, measuring the effectiveness of Facebook as a
learning tool has become an area of interest to academicians and
researchers. Therefore, this study tried to integrate and propose new
theoretical and empirical evidences by linking the western idea of
adopting Facebook as an alternative learning platform from a
Malaysian perspective. This study, thus, aimed to fill a gap by being
among the pioneering research that tries to study the effectiveness of
adopting Facebook as a learning platform across other cultural
settings, namely Malaysia. Structural equation modeling was
employed for data analysis and hypothesis testing. This study finding
has provided some insights that would likely affect students’
awareness towards using Facebook as an alternative learning
platform in the Malaysian higher learning institutions. At the end,
future direction is proposed.
Abstract: One of the key aspects of power quality improvement
in power system is the mitigation of voltage sags/swells and flicker.
Custom power devices have been known as the best tools for voltage
disturbances mitigation as well as reactive power compensation.
Dynamic Voltage Restorer (DVR) which is the most efficient and
effective modern custom power device can provide the most
commercial solution to solve several problems of power quality in
distribution networks. This paper deals with analysis and simulation
technique of DVR based on instantaneous power theory which is a
quick control to detect signals. The main purpose of this work is to
remove three important disturbances including voltage sags/swells
and flicker. Simulation of the proposed method was carried out on
two sample systems by using Matlab software environment and the
results of simulation show that the proposed method is able to
provide desirable power quality in the presence of wide range of
disturbances.
Abstract: This paper presents an optimal broadcast algorithm
for the hypercube networks. The main focus of the paper is the
effectiveness of the algorithm in the presence of many node faults.
For the optimal solution, our algorithm builds with spanning tree
connecting the all nodes of the networks, through which messages
are propagated from source node to remaining nodes. At any given
time, maximum n − 1 nodes may fail due to crashing. We show
that the hypercube networks are strongly fault-tolerant. Simulation
results analyze to accomplish algorithm characteristics under many
node faults. We have compared our simulation results between our
proposed method and the Fu’s method. Fu’s approach cannot tolerate
n − 1 faulty nodes in the worst case, but our approach can tolerate
n − 1 faulty nodes.
Abstract: Behavioral aspects of experience such as will power
are rarely subjected to quantitative study owing to the numerous
complexities involved. Will is a phenomenon that has puzzled
humanity for a long time. It is a belief that will power of an individual
affects the success achieved by them in life. It is also thought that a
person endowed with great will power can overcome even the most
crippling setbacks in life while a person with a weak will cannot make
the most of life even the greatest assets. This study is an attempt
to subject the phenomena of will to the test of an artificial neural
network through a computational model. The claim being tested is
that will power of an individual largely determines success achieved
in life. It is proposed that data pertaining to success of individuals
be obtained from an experiment and the phenomenon of will be
incorporated into the model, through data generated recursively using
a relation between will and success characteristic to the model.
An artificial neural network trained using part of the data, could
subsequently be used to make predictions regarding data points in
the rest of the model. The procedure would be tried for different
models and the model where the networks predictions are found to
be in greatest agreement with the data would be selected; and used
for studying the relation between success and will.
Abstract: The research of juice flavor forecasting has become
more important in China. Due to the fast economic growth in China,
many different kinds of juices have been introduced to the market. If a
beverage company can understand their customers’ preference well,
the juice can be served more attractive. Thus, this study intends to
introducing the basic theory and computing process of grapes juice
flavor forecasting based on support vector regression (SVR). Applying
SVR, BPN, and LR to forecast the flavor of grapes juice in real data
shows that SVR is more suitable and effective at predicting
performance.
Abstract: Wavelength Division Multiplexing (WDM) is the dominant transport technology used in numerous high capacity backbone networks, based on optical infrastructures. Given the importance of costs (CapEx and OpEx) associated to these networks, resource management is becoming increasingly important, especially how the optical circuits, called “lightpaths”, are routed throughout the network. This requires the use of efficient algorithms which provide routing strategies with the lowest cost. We focus on the lightpath routing and wavelength assignment problem, known as the RWA problem, while optimizing wavelength fragmentation over the network. Wavelength fragmentation poses a serious challenge for network operators since it leads to the misuse of the wavelength spectrum, and then to the refusal of new lightpath requests. In this paper, we first establish a new Integer Linear Program (ILP) for the problem based on a node-link formulation. This formulation is based on a multilayer approach where the original network is decomposed into several network layers, each corresponding to a wavelength. Furthermore, we propose an efficient heuristic for the problem based on a greedy algorithm followed by a post-treatment procedure. The obtained results show that the optimal solution is often reached. We also compare our results with those of other RWA heuristic methods
Abstract: In this paper static scheme of under-frequency based load shedding is considered for chemical and petrochemical industries with islanded distribution networks relying heavily on the primary commodity to ensure minimum production loss, plant downtime or critical equipment shutdown. A simplistic methodology is proposed for in-house implementation of this scheme using underfrequency relays and a step by step guide is provided including the techniques to calculate maximum percentage overloads, frequency decay rates, time based frequency response and frequency based time response of the system. Case study of FFL electrical system is utilized, presenting the actual system parameters and employed load shedding settings following the similar series of steps. The arbitrary settings are then verified for worst overload conditions (loss of a generation source in this case) and comprehensive system response is then investigated.
Abstract: This paper is concerned with knowledge representation
and extraction of fuzzy if-then rules using Interval Type-2
Context-based Fuzzy C-Means clustering (IT2-CFCM) with the aid of
fuzzy granulation. This proposed clustering algorithm is based on
information granulation in the form of IT2 based Fuzzy C-Means
(IT2-FCM) clustering and estimates the cluster centers by preserving
the homogeneity between the clustered patterns from the IT2 contexts
produced in the output space. Furthermore, we can obtain the
automatic knowledge representation in the design of Radial Basis
Function Networks (RBFN), Linguistic Model (LM), and Adaptive
Neuro-Fuzzy Networks (ANFN) from the numerical input-output data
pairs. We shall focus on a design of ANFN in this paper. The
experimental results on an estimation problem of energy performance
reveal that the proposed method showed a good knowledge
representation and performance in comparison with the previous
works.
Abstract: This paper presents a novel integrated hybrid
approach for fault diagnosis (FD) of nonlinear systems. Unlike most
FD techniques, the proposed solution simultaneously accomplishes
fault detection, isolation, and identification (FDII) within a unified
diagnostic module. At the core of this solution is a bank of adaptive
neural parameter estimators (NPE) associated with a set of singleparameter
fault models. The NPEs continuously estimate unknown
fault parameters (FP) that are indicators of faults in the system. Two
NPE structures including series-parallel and parallel are developed
with their exclusive set of desirable attributes. The parallel scheme is
extremely robust to measurement noise and possesses a simpler, yet
more solid, fault isolation logic. On the contrary, the series-parallel
scheme displays short FD delays and is robust to closed-loop system
transients due to changes in control commands. Finally, a fault
tolerant observer (FTO) is designed to extend the capability of the
NPEs to systems with partial-state measurement.
Abstract: Our globalizing world has become almost a small
village and everyone can access any information at any time.
Everyone lets each other know who does whatever in which place.
We can learn which social events occur in which place in the world.
From the perspective of education, the course notes that a lecturer use
in lessons in a university in any state of America can be examined by
a student studying in a city of Africa or the Far East. This dizzying
communication we have mentioned happened thanks to fast
developments in computer and internet technologies. While these
developments occur in the world, Turkey that has a very large young
population and whose electronic infrastructure rapidly improves has
also been affected by these developments. Nowadays, mobile devices
have become common and thus, it causes to increase data traffic in
social networks. This study was carried out on students in the
different age groups in Selcuk University Vocational School of
Technical Sciences, the Department of Computer Technology.
Students’ opinions about the use of internet and social media were
obtained. The features such as using the Internet and social media
skills, purposes, operating frequency, accessing facilities and tools,
social life and effects on vocational education and so forth were
explored. The positive effects and negative effects of both internet
and social media use on the students in this department and findings
are evaluated from different perspectives and results are obtained. In
addition, relations and differences were found out statistically.
Abstract: The aim of this paper is to present the optimization
methodology developed in the frame of a Coastal Transport
Information System. The system will be used for the effective design
of coastal transportation lines and incorporates subsystems that
implement models, tools and techniques that may support the design
of improved networks. The role of the optimization and decision
subsystem is to provide the user with better and optimal scenarios
that will best fulfill any constrains, goals or requirements posed. The
complexity of the problem and the large number of parameters and
objectives involved led to the adoption of an evolutionary method
(Genetic Algorithms). The problem model and the subsystem
structure are presented in detail, and, its support for simulation is also
discussed.
Abstract: In this paper, we introduced a gradient-based inverse
solver to obtain the missing boundary conditions based on the
readings of internal thermocouples. The results show that the method
is very sensitive to measurement errors, and becomes unstable when
small time steps are used. The artificial neural networks are shown to
be capable of capturing the whole thermal history on the run-out
table, but are not very effective in restoring the detailed behavior of
the boundary conditions. Also, they behave poorly in nonlinear cases
and where the boundary condition profile is different.
GA and PSO are more effective in finding a detailed
representation of the time-varying boundary conditions, as well as in
nonlinear cases. However, their convergence takes longer. A
variation of the basic PSO, called CRPSO, showed the best
performance among the three versions. Also, PSO proved to be
effective in handling noisy data, especially when its performance
parameters were tuned. An increase in the self-confidence parameter
was also found to be effective, as it increased the global search
capabilities of the algorithm. RPSO was the most effective variation
in dealing with noise, closely followed by CRPSO. The latter
variation is recommended for inverse heat conduction problems, as it
combines the efficiency and effectiveness required by these
problems.
Abstract: A model was constructed to predict the amount of
solar radiation that will make contact with the surface of the earth in
a given location an hour into the future. This project was supported
by the Southern Company to determine at what specific times during
a given day of the year solar panels could be relied upon to produce
energy in sufficient quantities. Due to their ability as universal
function approximators, an artificial neural network was used to
estimate the nonlinear pattern of solar radiation, which utilized
measurements of weather conditions collected at the Griffin, Georgia
weather station as inputs. A number of network configurations and
training strategies were utilized, though a multilayer perceptron with
a variety of hidden nodes trained with the resilient propagation
algorithm consistently yielded the most accurate predictions. In
addition, a modeled direct normal irradiance field and adjacent
weather station data were used to bolster prediction accuracy. In later
trials, the solar radiation field was preprocessed with a discrete
wavelet transform with the aim of removing noise from the
measurements. The current model provides predictions of solar
radiation with a mean square error of 0.0042, though ongoing efforts
are being made to further improve the model’s accuracy.
Abstract: An artificial neural network is a mathematical model
inspired by biological neural networks. There are several kinds of
neural networks and they are widely used in many areas, such as:
prediction, detection, and classification. Meanwhile, in day to day life,
people always have to make many difficult decisions. For example,
the coach of a soccer club has to decide which offensive player
to be selected to play in a certain game. This work describes a
novel Neural Network using a combination of the General Regression
Neural Network and the Probabilistic Neural Networks to help a
soccer coach make an informed decision.
Abstract: Wireless Sensor Networks (WSNs) have wide variety
of applications and provide limitless future potentials. Nodes in
WSNs are prone to failure due to energy depletion, hardware failure,
communication link errors, malicious attacks, and so on. Therefore,
fault tolerance is one of the critical issues in WSNs. We study how
fault tolerance is addressed in different applications of WSNs. Fault
tolerant routing is a critical task for sensor networks operating in
dynamic environments. Many routing, power management, and data
dissemination protocols have been specifically designed for WSNs
where energy awareness is an essential design issue. The focus,
however, has been given to the routing protocols which might differ
depending on the application and network architecture.
Abstract: Today’s VLSI networks demands for high speed. And
in this work the compact form mathematical model for current mode
signalling in VLSI interconnects is presented.RLC interconnect line
is modelled using characteristic impedance of transmission line and
inductive effect. The on-chip inductance effect is dominant at lower
technology node is emulated into an equivalent resistance. First order
transfer function is designed using finite difference equation, Laplace
transform and by applying the boundary conditions at the source and
load termination. It has been observed that the dominant pole
determines system response and delay in the proposed model. The
novel proposed current mode model shows superior performance as
compared to voltage mode signalling. Analysis shows that current
mode signalling in VLSI interconnects provides 2.8 times better
delay performance than voltage mode. Secondly the damping factor
of a lumped RLC circuit is shown to be a useful figure of merit.
Abstract: Nature is the immense gifted source for solving
complex problems. It always helps to find the optimal solution to
solve the problem. Mobile Ad Hoc NETwork (MANET) is a wide
research area of networks which has set of independent nodes. The
characteristics involved in MANET’s are Dynamic, does not depend
on any fixed infrastructure or centralized networks, High mobility.
The Bio-Inspired algorithms are mimics the nature for solving
optimization problems opening a new era in MANET. The typical
Swarm Intelligence (SI) algorithms are Ant Colony Optimization
(ACO), Artificial Bee Colony (ABC), Particle Swarm Optimization
(PSO), Modified Termite Algorithm, Bat Algorithm (BA), Wolf
Search Algorithm (WSA) and so on. This work mainly concentrated
on nature of MANET and behavior of nodes. Also it analyses various
performance metrics such as throughput, QoS and End-to-End delay
etc.
Abstract: Future mobile networks following 5th generation will
be characterized by one thousand times higher gains in capacity;
connections for at least one hundred billion devices; user experience
capable of extremely low latency and response times. To be close to
the capacity requirements and higher reliability, advanced
technologies have been studied, such as multiple connectivity, small
cell enhancement, heterogeneous networking, and advanced
interference and mobility management. This paper is focused on the
multiple connectivity in heterogeneous cellular networks. We
investigate the performance of coverage and user throughput in several
deployment scenarios. Using the stochastic geometry approach, the
SINR distributions and the coverage probabilities are derived in case
of dual connection. Also, to compare the user throughput enhancement
among the deployment scenarios, we calculate the spectral efficiency
and discuss our results.
Abstract: The Cone Penetration Test (CPT) is a common in-situ
test which generally investigates a much greater volume of soil more
quickly than possible from sampling and laboratory tests. Therefore,
it has the potential to realize both cost savings and assessment of soil
properties rapidly and continuously. The principle objective of this
paper is to demonstrate the feasibility and efficiency of using
artificial neural networks (ANNs) to predict the soil angle of internal
friction (Φ) and the soil modulus of elasticity (E) from CPT results
considering the uncertainties and non-linearities of the soil. In
addition, ANNs are used to study the influence of different
parameters and recommend which parameters should be included as
input parameters to improve the prediction. Neural networks discover
relationships in the input data sets through the iterative presentation
of the data and intrinsic mapping characteristics of neural topologies.
General Regression Neural Network (GRNN) is one of the powerful
neural network architectures which is utilized in this study. A large
amount of field and experimental data including CPT results, plate
load tests, direct shear box, grain size distribution and calculated data
of overburden pressure was obtained from a large project in the
United Arab Emirates. This data was used for the training and the
validation of the neural network. A comparison was made between
the obtained results from the ANN's approach, and some common
traditional correlations that predict Φ and E from CPT results with
respect to the actual results of the collected data. The results show
that the ANN is a very powerful tool. Very good agreement was
obtained between estimated results from ANN and actual measured
results with comparison to other correlations available in the
literature. The study recommends some easily available parameters
that should be included in the estimation of the soil properties to
improve the prediction models. It is shown that the use of friction
ration in the estimation of Φ and the use of fines content in the
estimation of E considerable improve the prediction models.