Abstract: The quality of short term load forecasting can improve the efficiency of planning and operation of electric utilities. Artificial Neural Networks (ANNs) are employed for nonlinear short term load forecasting owing to their powerful nonlinear mapping capabilities. At present, there is no systematic methodology for optimal design and training of an artificial neural network. One has often to resort to the trial and error approach. This paper describes the process of developing three layer feed-forward large neural networks for short-term load forecasting and then presents a heuristic search algorithm for performing an important task of this process, i.e. optimal networks structure design. Particle Swarm Optimization (PSO) is used to develop the optimum large neural network structure and connecting weights for one-day ahead electric load forecasting problem. PSO is a novel random optimization method based on swarm intelligence, which has more powerful ability of global optimization. Employing PSO algorithms on the design and training of ANNs allows the ANN architecture and parameters to be easily optimized. The proposed method is applied to STLF of the local utility. Data are clustered due to the differences in their characteristics. Special days are extracted from the normal training sets and handled separately. In this way, a solution is provided for all load types, including working days and weekends and special days. The experimental results show that the proposed method optimized by PSO can quicken the learning speed of the network and improve the forecasting precision compared with the conventional Back Propagation (BP) method. Moreover, it is not only simple to calculate, but also practical and effective. Also, it provides a greater degree of accuracy in many cases and gives lower percent errors all the time for STLF problem compared to BP method. Thus, it can be applied to automatically design an optimal load forecaster based on historical data.
Abstract: The main problem for recognition of handwritten Persian digits using Neural Network is to extract an appropriate feature vector from image matrix. In this research an asymmetrical segmentation pattern is proposed to obtain the feature vector. This pattern can be adjusted as an optimum model thanks to its one degree of freedom as a control point. Since any chosen algorithm depends on digit identity, a Neural Network is used to prevail over this dependence. Inputs of this Network are the moment of inertia and the center of gravity which do not depend on digit identity. Recognizing the digit is carried out using another Neural Network. Simulation results indicate the high recognition rate of 97.6% for new introduced pattern in comparison to the previous models for recognition of digits.
Abstract: Purpose of this work is the development of an
automatic classification system which could be useful for radiologists
in the investigation of breast cancer. The software has been designed
in the framework of the MAGIC-5 collaboration.
In the automatic classification system the suspicious regions with
high probability to include a lesion are extracted from the image as
regions of interest (ROIs). Each ROI is characterized by some
features based on morphological lesion differences.
Some classifiers as a Feed Forward Neural Network, a K-Nearest
Neighbours and a Support Vector Machine are used to distinguish the
pathological records from the healthy ones.
The results obtained in terms of sensitivity (percentage of
pathological ROIs correctly classified) and specificity (percentage of
non-pathological ROIs correctly classified) will be presented through
the Receive Operating Characteristic curve (ROC). In particular the
best performances are 88% ± 1 of area under ROC curve obtained
with the Feed Forward Neural Network.
Abstract: Today automobile and aerospace industries realise Laser Beam Welding for a clean and non contact source of heating and fusion for joining of sheets. The welding performance is mainly based on by the laser welding parameters. Some concepts related to Artificial Neural Networks and how can be applied to model weld bead geometry and mechanical properties in terms of equipment parameters are reported in order to evaluate the accuracy and compare it with traditional modeling schemes. This review reveals the output features of Titanium and Aluminium weld bead geometry and mechanical properties such as ultimate tensile strength, yield strength, elongation and reduction of the area of the weld using Artificial Neural Network.
Abstract: A new approach to promote the generalization ability
of neural networks is presented. It is based on the point of view of
fuzzy theory. This approach is implemented through shrinking or
magnifying the input vector, thereby reducing the difference between
training set and testing set. It is called “shrinking-magnifying
approach" (SMA). At the same time, a new algorithm; α-algorithm is
presented to find out the appropriate shrinking-magnifying-factor
(SMF) α and obtain better generalization ability of neural networks.
Quite a few simulation experiments serve to study the effect of SMA
and α-algorithm. The experiment results are discussed in detail, and
the function principle of SMA is analyzed in theory. The results of
experiments and analyses show that the new approach is not only
simpler and easier, but also is very effective to many neural networks
and many classification problems. In our experiments, the proportions
promoting the generalization ability of neural networks have even
reached 90%.
Abstract: Diabetes Mellitus is a chronic metabolic disorder, where the improper management of the blood glucose level in the diabetic patients will lead to the risk of heart attack, kidney disease and renal failure. This paper attempts to enhance the diagnostic accuracy of the advancing blood glucose levels of the diabetic patients, by combining principal component analysis and wavelet neural network. The proposed system makes separate blood glucose prediction in the morning, afternoon, evening and night intervals, using dataset from one patient covering a period of 77 days. Comparisons of the diagnostic accuracy with other neural network models, which use the same dataset are made. The comparison results showed overall improved accuracy, which indicates the effectiveness of this proposed system.
Abstract: This paper proposes a neural network weights and
topology optimization using genetic evolution and the
backpropagation training algorithm. The proposed crossover and
mutation operators aims to adapt the networks architectures and
weights during the evolution process. Through a specific inheritance
procedure, the weights are transmitted from the parents to their
offsprings, which allows re-exploitation of the already trained
networks and hence the acceleration of the global convergence of the
algorithm. In the preprocessing phase, a new feature extraction
method is proposed based on Legendre moments with the Maximum
entropy principle MEP as a selection criterion. This allows a global
search space reduction in the design of the networks. The proposed
method has been applied and tested on the well known MNIST
database of handwritten digits.
Abstract: In this paper, an automatic determination algorithm for nuclear magnetic resonance (NMR) spectra of the metabolites in the living body by magnetic resonance spectroscopy (MRS) without human intervention or complicated calculations is presented. In such method, the problem of NMR spectrum determination is transformed into the determination of the parameters of a mathematical model of the NMR signal. To calculate these parameters efficiently, a new model called modified Hopfield neural network is designed. The main achievement of this paper over the work in literature [30] is that the speed of the modified Hopfield neural network is accelerated. This is done by applying cross correlation in the frequency domain between the input values and the input weights. The modified Hopfield neural network can accomplish complex dignals perfectly with out any additinal computation steps. This is a valuable advantage as NMR signals are complex-valued. In addition, a technique called “modified sequential extension of section (MSES)" that takes into account the damping rate of the NMR signal is developed to be faster than that presented in [30]. Simulation results show that the calculation precision of the spectrum improves when MSES is used along with the neural network. Furthermore, MSES is found to reduce the local minimum problem in Hopfield neural networks. Moreover, the performance of the proposed method is evaluated and there is no effect on the performance of calculations when using the modified Hopfield neural networks.
Abstract: In this paper, an ultrasonic technique is proposed to
predict oil content in a fresh palm fruit. This is accomplished by
measuring the attenuation based on ultrasonic transmission mode.
Several palm fruit samples with known oil content by Soxhlet
extraction (ISO9001:2008) were tested with our ultrasonic
measurement. Amplitude attenuation data results for all palm samples
were collected. The Feedforward Neural Networks (FNNs) are
applied to predict the oil content for the samples. The Root Mean
Square Error (RMSE) and Mean Absolute Error (MAE) of the FNN
model for predicting oil content percentage are 7.6186 and 5.2287
with the correlation coefficient (R) of 0.9193.
Abstract: This paper introduces two decoders for binary linear
codes based on Metaheuristics. The first one uses a genetic algorithm
and the second is based on a combination genetic algorithm with
a feed forward neural network. The decoder based on the genetic
algorithms (DAG) applied to BCH and convolutional codes give good
performances compared to Chase-2 and Viterbi algorithm respectively
and reach the performances of the OSD-3 for some Residue
Quadratic (RQ) codes. This algorithm is less complex for linear
block codes of large block length; furthermore their performances
can be improved by tuning the decoder-s parameters, in particular the
number of individuals by population and the number of generations.
In the second algorithm, the search space, in contrast to DAG which
was limited to the code word space, now covers the whole binary
vector space. It tries to elude a great number of coding operations
by using a neural network. This reduces greatly the complexity of
the decoder while maintaining comparable performances.
Abstract: Globalization and therefore increasing tight competition among companies, have resulted to increase the importance of making well-timed decision. Devising and employing effective strategies, that are flexible and adaptive to changing market, stand a greater chance of being effective in the long-term. In other side, a clear focus on managing the entire product lifecycle has emerged as critical areas for investment. Therefore, applying wellorganized tools to employ past experience in new case, helps to make proper and managerial decisions. Case based reasoning (CBR) is based on a means of solving a new problem by using or adapting solutions to old problems. In this paper, an adapted CBR model with k-nearest neighbor (K-NN) is employed to provide suggestions for better decision making which are adopted for a given product in the middle of life phase. The set of solutions are weighted by CBR in the principle of group decision making. Wrapper approach of genetic algorithm is employed to generate optimal feature subsets. The dataset of the department store, including various products which are collected among two years, have been used. K-fold approach is used to evaluate the classification accuracy rate. Empirical results are compared with classical case based reasoning algorithm which has no special process for feature selection, CBR-PCA algorithm based on filter approach feature selection, and Artificial Neural Network. The results indicate that the predictive performance of the model, compare with two CBR algorithms, in specific case is more effective.
Abstract: In this paper back-propagation artificial neural
network (BPANN) is employed to predict the limiting drawing ratio
(LDR) of the deep drawing process. To prepare a training set for
BPANN, some finite element simulations were carried out. die and
punch radius, die arc radius, friction coefficient, thickness, yield
strength of sheet and strain hardening exponent were used as the
input data and the LDR as the specified output used in the training of
neural network. As a result of the specified parameters, the program
will be able to estimate the LDR for any new given condition.
Comparing FEM and BPANN results, an acceptable correlation was
found.
Abstract: The myoelectric signal (MES) is one of the Biosignals
utilized in helping humans to control equipments. Recent approaches
in MES classification to control prosthetic devices employing pattern
recognition techniques revealed two problems, first, the classification
performance of the system starts degrading when the number of
motion classes to be classified increases, second, in order to solve the
first problem, additional complicated methods were utilized which
increase the computational cost of a multifunction myoelectric
control system. In an effort to solve these problems and to achieve a
feasible design for real time implementation with high overall
accuracy, this paper presents a new method for feature extraction in
MES recognition systems. The method works by extracting features
using Wavelet Packet Transform (WPT) applied on the MES from
multiple channels, and then employs Fuzzy c-means (FCM)
algorithm to generate a measure that judges on features suitability for
classification. Finally, Principle Component Analysis (PCA) is
utilized to reduce the size of the data before computing the
classification accuracy with a multilayer perceptron neural network.
The proposed system produces powerful classification results (99%
accuracy) by using only a small portion of the original feature set.
Abstract: The paper describes a self supervised parallel self organizing neural network (PSONN) architecture for true color image segmentation. The proposed architecture is a parallel extension of the standard single self organizing neural network architecture (SONN) and comprises an input (source) layer of image information, three single self organizing neural network architectures for segmentation of the different primary color components in a color image scene and one final output (sink) layer for fusion of the segmented color component images. Responses to the different shades of color components are induced in each of the three single network architectures (meant for component level processing) by applying a multilevel version of the characteristic activation function, which maps the input color information into different shades of color components, thereby yielding a processed component color image segmented on the basis of the different shades of component colors. The number of target classes in the segmented image corresponds to the number of levels in the multilevel activation function. Since the multilevel version of the activation function exhibits several subnormal responses to the input color image scene information, the system errors of the three component network architectures are computed from some subnormal linear index of fuzziness of the component color image scenes at the individual level. Several multilevel activation functions are employed for segmentation of the input color image scene using the proposed network architecture. Results of the application of the multilevel activation functions to the PSONN architecture are reported on three real life true color images. The results are substantiated empirically with the correlation coefficients between the segmented images and the original images.
Abstract: In this paper, a novel approach for robust trajectory tracking of induction motor drive is presented. By combining variable structure systems theory with fuzzy logic concept and neural network techniques, a new algorithm is developed. Fuzzy logic was used for the adaptation of the learning algorithm to improve the robustness of learning and operating of the neural network. The developed control algorithm is robust to parameter variations and external influences. It also assures precise trajectory tracking with the prescribed dynamics. The algorithm was verified by simulation and the results obtained demonstrate the effectiveness of the designed controller of induction motor drives which considered as highly non linear dynamic complex systems and variable characteristics over the operating conditions.
Abstract: In this paper, an approach to reduce the computation steps required by fast neural networksfor the searching process is presented. The principle ofdivide and conquer strategy is applied through imagedecomposition. Each image is divided into small in sizesub-images and then each one is tested separately usinga fast neural network. The operation of fast neuralnetworks based on applying cross correlation in thefrequency domain between the input image and theweights of the hidden neurons. Compared toconventional and fast neural networks, experimentalresults show that a speed up ratio is achieved whenapplying this technique to locate human facesautomatically in cluttered scenes. Furthermore, fasterface detection is obtained by using parallel processingtechniques to test the resulting sub-images at the sametime using the same number of fast neural networks. Incontrast to using only fast neural networks, the speed upratio is increased with the size of the input image whenusing fast neural networks and image decomposition.
Abstract: During signal transmission, the combined effect of the
transmitter filter, the transmission medium, and additive white
Gaussian noise (AWGN) are included in the channel which distort
and add noise to the signal. This causes the well defined signal
constellation to spread causing errors in bit detection. A compact pi
neural network with minimum number of nodes is proposed. The
replacement of summation at each node by multiplication results in
more powerful mapping. The resultant pi network is tested on six
different channels.
Abstract: The various applications of VLSI circuits in highperformance
computing, telecommunications, and consumer
electronics has been expanding progressively, and at a very hasty
pace. This paper describes a new model for partitioning a circuit
using DBSCAN and fuzzy ARTMAP neural network. The first step
is concerned with feature extraction, where we had make use
DBSCAN algorithm. The second step is the classification and is
composed of a fuzzy ARTMAP neural network. The performance of
both approaches is compared using benchmark data provided by
MCNC standard cell placement benchmark netlists. Analysis of the
investigational results proved that the fuzzy ARTMAP with
DBSCAN model achieves greater performance then only fuzzy
ARTMAP in recognizing sub-circuits with lowest amount of
interconnections between them The recognition rate using fuzzy
ARTMAP with DBSCAN is 97.7% compared to only fuzzy
ARTMAP.
Abstract: Ability of accurate and reliable location estimation in
indoor environment is the key issue in developing great number of
context aware applications and Location Based Services (LBS).
Today, the most viable solution for localization is the Received
Signal Strength (RSS) fingerprinting based approach using wireless
local area network (WLAN). This paper presents two RSS
fingerprinting based approaches – first we employ widely used
WLAN based positioning as a reference system and then investigate
the possibility of using GSM signals for positioning. To compare
them, we developed a positioning system in real world environment,
where realistic RSS measurements were collected. Multi-Layer
Perceptron (MLP) neural network was used as the approximation
function that maps RSS fingerprints and locations. Experimental
results indicate advantage of WLAN based approach in the sense of
lower localization error compared to GSM based approach, but GSM
signal coverage by far outreaches WLAN coverage and for some
LBS services requiring less precise accuracy our results indicate that
GSM positioning can also be a viable solution.
Abstract: The backpropagation algorithm in general employs quadratic error function. In fact, most of the problems that involve minimization employ the Quadratic error function. With alternative error functions the performance of the optimization scheme can be improved. The new error functions help in suppressing the ill-effects of the outliers and have shown good performance to noise. In this paper we have tried to evaluate and compare the relative performance of complex valued neural network using different error functions. During first simulation for complex XOR gate it is observed that some error functions like Absolute error, Cauchy error function can replace Quadratic error function. In the second simulation it is observed that for some error functions the performance of the complex valued neural network depends on the architecture of the network whereas with few other error functions convergence speed of the network is independent of architecture of the neural network.