Search

The CORBA based middlewares are appropriate for Telematics system which is an important and practical infrastructure for ubiquitous convergence. And fault tolerance is a first priority for real time characteristics of transport information provided by Telematics systems. There are many methods to enhance the fault tolerance of CORBA based systems. Among these methods, this paper proposes the one of enabling seamless information services where systems based on the CORBA suffer from object faults which occur processing real time transport information. Namely, this paper observes a method to deal efficiently with object faults occurring in 3 tier architecture environments. It is possible to replicate objects as a method to enhance fault tolerance in preparation against object faults. This paper presents a method of enhancing Telematics fault tolerance and maintaining service continuity by continuing to operate systems until faults of the Fault Tolerant (FT)-CORBA protocol itself is recovered. The proposed agent enhances Telematics performance by adjusting the minimum number of object replicas for real time transport information services during system operation. The evaluation is performed by comparing costs for system fault tolerance which consist of communication traffic and S/W complexities. [ABSTRACT FROM AUTHOR]

Intelligent Agents have originated a lot discussion about what they are, and how they are a different from general programs. We describe in this paper a new paradigm for intelligent agents, This paradigm helped us deal with failures in an independent and efficient way. We proposed these types of agents to treat the system in a hierarchical way. The Agent Node is also described. A new method to visualize fault tolerant system (FTS) is proposed, in this paper witch the incorporation of intelligent agents, which as they grow and specialized create the Agent System (MAS).One is the communications diagrams of the each of the agents, described in diagrams of transaction of states. [ABSTRACT FROM AUTHOR]

Single input Fuzzy TPE systems, as proposed by Sudkamp and Hammel, allow a more efficient computational inference of a single input fuzzy rule base. This paper shows that it is possible to generalize single input Fuzzy TPE systems to 2-input systems, therefore extending its range of possible applications. The paper presents the details and proof of the extension validity, and shows benchmark results comparing the 2-input Fuzzy TPE with classical inference systems. [ABSTRACT FROM AUTHOR]

The Particle Swarm Optimization (PSO) and the Genetic Algorithms (GA) have been used successfully in solving problems of optimization with continuous and combinatorial search spaces. In this paper the results of the application of PSO and GAs for the optimization of mathematical functions is presented. These two methodologies have been implemented with the goal of making a comparison of their performance in solving complex optimization problems. This paper describes a comparison between a GA and PSO for the optimization of a complex mathematical function. [ABSTRACT FROM AUTHOR]

Process Planning activities are significantly based on experience and technical skill. In spite of the great efforts made for planning automation, this activity continues being made in manual form. Process Planning activities are significantly based on experience and technical skills. The advent of the CAM systems (Computer Aided Manufacturing) has partially close the gap left between the Automated Design and Manufacture. Meanwhile, a great dose of manual work still exists and investigation in this area is still necessary. This paper presents the application of a multi objective genetic algorithm for the definition of the optimal cutting parameters. The objective functions consider the production rate and production cost in turning operations. The obtained Pareto front is compared to high efficiency cutting range. This paper also describes one application of the developed mechanism using an example. [ABSTRACT FROM AUTHOR]

We propose a new method for handwriting recognition that utilizes geometric features of letters. The paper deals with recognition of isolated handwritten characters using an artificial neural network. The characters are written on a regular sheet of paper using a pen, and then they are captured optically by a scanner and processed to a binary image which is analyzed by a computer. In this paper we present a new method for off-line handwriting recognition and also describe our research and tests performed on the neural network. [ABSTRACT FROM AUTHOR]

This paper proposes a new method for fuzzy rule extraction from trained support vector machines (SVMs) for multi-class problems. SVMs have been applied to a wide variety of application. However, SVMs are considered "black box models", where no interpretation about the input-output mapping is provided. Some methods to reduce this limitation have already been proposed, however, they are restricted to binary classification problems and to the extraction of symbolic rules with intervals or functions in their antecedents. Hence, to improve the interpretability of the generated rules, this paper presents a new model for extracting fuzzy rules from a trained SVM. Moreover, the proposed model was developed for classification in multi-class problems. The generated fuzzy rules are presented in the format "IF x1 is C1 AND x2 is C2 AND ... AND xn is Cn, THEN x = (x1, x2,... xn) is of class A", where C1, C2..., Cn are fuzzy sets. The proposed method was evaluated in four benchmark databases (Bupa Liver Disorders, Wisconsin Breast Cancer, Iris and Wine). The results obtained demonstrate the capacity of the proposed method to generate a set of interpretable rules that explains the database and the influence of the input variables in the determination of the final class. [ABSTRACT FROM AUTHOR]

In this paper, an intelligent home environment with an adaptive type-2 fuzzy logic (AT2FL) method is developed. The considered system is mainly for elderly and disabled persons and catches the users' intention through recognizing their hand gestures instead of remote controller. However, during the recognition procedure, to distinguish between the command and ordinary gestures needs a very rigorous procedure, and to cope with that, the AT2FL approach is proposed in this paper. The presented algorithm is applied to control home appliances, and some experimental results will be shown to verify the feasibility and effectiveness. [ABSTRACT FROM AUTHOR]

We describe in this paper a new method for response integration in modular neural networks using type-2 fuzzy logic. The modular neural networks were used in human person recognition. Biometric authentication is used to achieve person recognition. Three biometric characteristics of the person are used: face, fingerprint, and voice. A modular neural network of three modules is used. Each module is a local expert on person recognition based on each of the biometric measures. The response integration method of the modular neural network has the goal of combining the responses of the modules to improve the recognition rate of the individual modules. We show in this paper results of a type-2 fuzzy approach for response integration that improves performance over type-1 fuzzy logic approaches. [ABSTRACT FROM AUTHOR]

Interactive digital media, specifically video games, have emerged as a dominant new economic, cultural and, recently, educational force. The prevalence of video games in society today has actually re-wired our brains and has made traditional learning less effective. Today's digital learning methods for young learners fail to engage audiences accustomed to interactive media. And as gaming environments become more complex, games may be more useful in providing alternative paths to hard to teach concepts. This paper introduces an intelligent Belief Desire Intention (BDI) agent architecture for a Non-Player Character that encourages and stimulates situated learning in an online Role-Playing Game. [ABSTRACT FROM AUTHOR]

Companies are pursuing a new approach through which overall functions within a supply chain, ranging from material purchase to production, distribution, and sales are designed, planned, and managed in an integrated way. The core functions among those functions are production planning and distribution planning. In this paper, we propose a multi-agent system and genetic algorithm that make it possible to integrate the production and distribution planning on a real-time basis in supply chain management. [ABSTRACT FROM AUTHOR]

Control charts are widely used to establish statistical process control (SPC) and maintain manufacturing processes in desired operating conditions. Effective use of a control chart requires finding the best parameters for the optimal operation of the chart, which is called design of control chart. The design of control chart involves the selection of three parameters namely, the sample size, the sampling interval, and the control limits coefficients. Conventional approaches to design control charts include complex mathematical models and associated optimization schemes. Also, Conventional approaches usually use point estimators of the input parameters which are not able to represent the true parameters sufficiently. The input parameters are suffered from ambiguity uncertainty, which can be effectively modeled by using fuzzy set theory. In this paper a fuzzy multi-objective model for economic-statistical design of X-bar control chart, in which the input parameters are expressed by fuzzy membership functions, is proposed. Moreover, an ANFIS approach is used to transform the complex fuzzy multi-objective model into a fuzzy rule-base. Next, a genetic algorithm (GA) has been developed to find the best design (input) parameters of control charts. Finally, the proposed approach is tested and validated by using a numerical example. [ABSTRACT FROM AUTHOR]

In this paper, a fuzzy model is proposed for supplier selection and to identify proper Supplier Development Programs (SDPs). A fuzzy rule base is used to calculate the utility value of a firm's managers to perform a special SDP. Implementation of the model is demonstrated by a numerical example. [ABSTRACT FROM AUTHOR]

We describe in this paper a new approach for pattern recognition using modular neural networks with a fuzzy logic method for response integration. We proposed a new architecture for modular neural networks for achieving pattern recognition in the particular case of human fingerprints. Also, the method for achieving response integration is based on the fuzzy Sugeno integral. Response integration is required to combine the outputs of all the modules in the modular network. We have applied the new approach for fingerprint recognition with a real database of fingerprints from students of our institution. [ABSTRACT FROM AUTHOR]

This paper presents a generalized method for tuning a fuzzy logic controller based on reinforcement learning in a dynamic environment. We extend the Generalized Approximate Reasoning-base Intelligent Controller (GARIC) model of Berenji and Khedkar to be able to work in presence of vagueness in states. Similar to GARIC, the proposed architecture, i.e., Generalized Reinforcement Learning Fuzzy Controller (GRLFC), has the self-tuning capability even when only a weak reinforcement signal such a binary failure signal is available. The proposed controller shows a better performance, regarding learning speed and robustness to changes in controlled system dynamics, than similar models even in the presence of uncertainty in states. [ABSTRACT FROM AUTHOR]

This paper describes the use of a Genetic Algorithm (GA) for the problem of Offline Point-to-Point Autonomous Mobile Robot Path Planning. The problem consist of generating "valid" paths or trajectories, for an Holonomic Robot to use to move from a starting position to a destination across a flat map of a terrain, represented by a two dimensional grid, with obstacles and dangerous ground that the Robot must evade. This means that the GA optimizes possible paths based on two criteria: length and difficulty. [ABSTRACT FROM AUTHOR]

In this paper, the robust stability analysis of a fuzzy vehicle lateral system with perturbed parameters is presented. Firstly, the fuzzy controller can be linearized by utilizing the describing function method with experiments. After the describing function is obtained, the stability analysis of the vehicle lateral control system with the variations of velocity and friction is then carried out by the use of parameter plane method. Afterward some limit cycle loci caused by the fuzzy controller can be easily pointed out in the parameter plane. Computer simulation shows the efficiency of this approach. [ABSTRACT FROM AUTHOR]

This paper presents an adaptive, support vector machine-based ECG signal processing and compression method. After a conventional pre-filtering step, the characteristic waves (QRS, P, T) from the ECG signal are localized. The following step contains a regressive model for waveform description in terms of model parameters. The gained information allows an iterative filtering in permanent concordance with the aimed processing manner. The structure of the algorithm allows real-time adaptation to the heart's state. Using these methods for one channel of the MIT-BIH database, the detection rate of QRS complexes is above 99.9%. The negative influence of various noise types, like 50/60 Hz power line, abrupt baseline shift or drift, and low sampling rate was almost completely eliminated. The vector support machine system allow a good balance between compressing and diagnostic performance and the obtained results can form a solid base for better data storage in clinical environment. [ABSTRACT FROM AUTHOR]

In this paper we integrate the self-adjustment environment advantages of reinforcement learning into the adaptive wavelet network controller. The novel approach is called adaptive wavelet reinforcement learning control, which uses wavelet to approximate a continuous Q-function, in order to obtain a optimal control policy. The simulations of applying the proposed method to an inverted pendulum and Pendubot system are performed to demonstrate the properties of the adaptive wavelet network controller. [ABSTRACT FROM AUTHOR]

There are different approximations to improve the performance and mathematical representation of a cellular neural networks to work with linearly nonseparable data as XOR. But the main goal is to work with problems that only can solved with universal machines such as the game of life. In this paper a new model of Polynomial Cellular Neural Networks that solves the game of life is presented with the learning design to compute the templates. [ABSTRACT FROM AUTHOR]

Ensemble Neural Networks are a learning paradigm where many neural networks are used together to solve a particular problem. In this paper, the relationship between the ensemble and its component neural networks is analyzed with the goal of creating of a set of nets for an ensemble with the use of a sampling-technique. This technique is such that each net in the ensemble is trained on a different sub-sample of the training data. [ABSTRACT FROM AUTHOR]

The selection of appropriate machines is one of the most crucial decisions for a manufacturing company to develop an efficient production environment. The aim of this study is to propose a fuzzy approach for selecting the best machine. This paper is based on a fuzzy extension of the TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) method. In this method, the ratings of various alternatives versus various subjective criteria and the weights of all criteria are assessed in linguistic variables represented by fuzzy numbers. Fuzzy numbers try to resolve the ambiguity of concepts that are associated with human being's judgments. To determine the order of the alternatives, closeness coefficient is defined by calculating the distances to the fuzzy positive ideal solution (FPIS) and fuzzy negative ideal solution (FNIS). By using fuzzy TOPSIS, uncertainty and vagueness from subjective perception and the experiences of decision maker can be effectively represented and reached to a more effective decision. [ABSTRACT FROM AUTHOR]

Resolution principle is the rule of inference at the basis of most procedures for automated reasoning both in classic logic and possibilistic logic. In these procedures, deduction starts by inferring new clauses by resolution, and goes on until the empty clause is generated or consistency of the set of clauses is proven, e.g. because no new clauses can be generated. Theorem proving using extension rule is a novel approach for reasoning in classic logic advanced by us recently. The key idea is to use the inverse of resolution and to use the inclusion-exclusion principle to circumvent the problem of space complexity. In this paper, we proposed a possibilistic extension of "extension rule", which we called possibilistic extension rule. We showed that it outperformed possibilistic resolution-based method in some cases. And it is potentially a complementary method to possibilistic resolution rule. [ABSTRACT FROM AUTHOR]

Location services are used in mobile ad hoc and hybrid networks either to locate the geographic position of a given node in the network or for locating a data item. One of the main usages of position location services is in location based routing algorithms. In particular, geographic routing protocols can route messages more efficiently to their destinations based on the destination node's geographic position, which is provided by a location service. In this paper, we propose an adaptive location service on the basis of fuzzy logic called FHLS (Fuzzy Hierarchical Location Service) for mobile ad hoc networks. The FHLS uses the adaptive location update scheme using the fuzzy logic on the basis of the mobility and the call preference of mobile nodes. The performance of the FHLS is to be evaluated by using a simulation, and compared with that of existing HLS scheme. [ABSTRACT FROM AUTHOR]

Fuzzy measures are used to describe interactions among given attributes towards a certain target. Given an observation as a function defined on the set of attributes, an integral of the function with respect to a given fuzzy measure is usually taken as an aggregation tool in information fusion. Various types of integrals represent different coordination modes of attributes. Thus, in case the coordination mode is unknown, due to the nonadditivity of the fuzzy measure, the integral value is uncertain and varies according to the type of the integral. A commonly used type of fuzzy measures are believe measures. The coordination uncertainty associated with belief measures in information fusion is discussed. Furthermore, the behavior of coordination uncertainty under the Dempster rule of combination is investigated in this paper. [ABSTRACT FROM AUTHOR]

The purpose of this paper is to present an effective framework PC-TOPSIS when using both Analytic Hierarchy Process (AHP) to address the weights of the criteria, and Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) extended to a fuzzy environmental for performing alternatives evaluation. Thus we could handle the decision problems in a more realist way. An example is provided to illustrate the application of the proposed PC-TOPSIS in a decision-making (DM) of selection of a cleaning system in a maintenance problem. [ABSTRACT FROM AUTHOR]

Because the geological events are neither homogeneous nor isotropic, the geological investigations are characterized by particularly high uncertainties. This paper presents a hybrid methodology for measuring of uncertainty in regional grade variability. In order to evaluate the fuzziness in grade values at ore deposit, point cumulative semimadogram (PCSM) measure and a metric distance have been employed. By using the experimental PCSMs and their linear models, measures of fuzziness have been carried out for each location. Finally, an uncertainty map, which defines the regional variation of the uncertainty in different categories, has been composed. [ABSTRACT FROM AUTHOR]

It plays an important role in sensory evaluation to optimize experiment design, i.e. using less number of tests to obtain available data as possible by intelligent technologies. This paper presents one method for optimizing experiment design based on learning automaton, and this method is applied in computer aided sensory evaluation (abbr. CASE). The validity of this method is showed by the result of experiment from CASE. [ABSTRACT FROM AUTHOR]

The paper deals with an improvement of implementation of Micro Hydro Power Plant. The main idea is to apply fuzzy control in such rather uncertain, ill defined and difficult problem as is in fact continuous power production in variable conditions: water supply, energy consumption, and different irregular consumers' loads. The aim is to demonstrate some advantages of fuzzy control able to resolve these problems borrowing human intuition, insight, and creative thinking inherent so far to human being only. Some results of the experiments are demonstrated on pilot MHPP, which now is in active stage of its exploitation. These experiments show a comparison between conventional microprocessor control and such with addition of fuzzy logic contours, which transform in fact the system in a hybrid control system. [ABSTRACT FROM AUTHOR]

In this paper we illustrate the LU representation of fuzzy numbers and present an LU-fuzzy calculator, in order to explain the use of the LU-fuzzy model and to show the advantage of the parametrization. The model can be applied either in the level-cut or in generalized LR frames. The hand-like fuzzy calculator has been developed for the MS-Windows platform and produces the basic fuzzy calculus: the arithmetic operations (scalar multiplication, addition, subtraction, multiplication, division) and the fuzzy extension of many univariate functions (exponential, logarithm, power with numeric or fuzzy exponent, sin, arcsin, cos, arccos, tan, arctan, square root, Gaussian, hyperbolic sinh, cosh, tanh and inverses, erf and erfc error functions, cumulative standard normal distribution). [ABSTRACT FROM AUTHOR]

This paper describes the development of a closed-loop attitude estimation system for determining attitude reference for vehicle dynamics using fuzzy inference and Genetic Algorithms (GAs). By recognizing the situation of dynamic condition via fuzzy inference process, each parameter of the estimator of the attitude estimation system is determined online adaptively under varying vehicle dynamics. For this solution scheme, fuzzy rules and reasoning method are consider based on the error signal of the gyro and accelerometer and the magnitude of dynamic motion, and the input gains of the fuzzy systems and the position of the membership function are optimized based on the GAs. Computer simulations based on the real test data of a vehicle are used in the study to assess the system performance with the proposed fuzzy-GAs estimation method. [ABSTRACT FROM AUTHOR]

In this paper we apply a new evolutive approach for solving the Set Covering Problem. This problem is a reasonably well known NP-complete optimization problem with many real world applications. We use a Cultural Evolutionary Architecture to maintain knowledge of Diversity and Fitness learned over each generation during the search process. Our results indicate that the approach is able to produce competitive results in compare with other approximation algorithms solving a portfolio of test problems taken from the ORLIB. [ABSTRACT FROM AUTHOR]

In this paper, we first present a learning algorithm for dynamic recurrent Elman neural networks based on a dissimilation particle swarm optimization. The proposed algorithm computes concurrently both the evolution of network structure, weights, initial inputs of the context units and self-feedback coefficient of the modified Elman network. Thereafter, we introduce and discuss a novel control method based on the proposed algorithm. More specifically, a dynamic identifier is constructed to perform speed identification and a controller is designed to perform speed control for Ultrasonic Motors (USM). Numerical experiments show that the novel identifier and controller based on the proposed algorithm can both achieve higher convergence precision and speed. In particular, our experiments show that the identifier can approximate the USM's nonlinear input-output mapping accurately. The effectiveness of the controller is verified using different kinds of speeds of constant and sinusoidal types. Besides, a preliminary examination on a randomly perturbation also shows the robust characteristics of the two proposed models. [ABSTRACT FROM AUTHOR]

Genetic fuzzy rule selection is an effective approach to the design of accurate and interpretable fuzzy rule-based classifiers. It tries to minimize the complexity of fuzzy rule-based classifiers while maximizing their accuracy by selecting only a small number of fuzzy rules from a large number of candidate rules. One important issue in genetic fuzzy rule selection is the prescreening of candidate rules. If we use too many candidate rules, a large computation load is required to search for good rule sets. On the other hand, good rule sets will not be found when promising fuzzy rules are not included in the candidate rule set. It is essential for the success of genetic fuzzy rule selection to use a tractable number of promising fuzzy rules as candidate rules. In this paper, we propose an idea of using Pareto-optimal and near Pareto-optimal fuzzy rules as candidate rules in genetic fuzzy rule selection. Pareto-optimality is defined by two well-known data mining criteria: support and confidence. To extract not only Pareto-optimal but also near Pareto-optimal fuzzy rules, we modify Pareto dominance using a dominance margin ε. Through computational experiments, we examine the effect of the proposed idea on multiobjective genetic fuzzy rule selection. [ABSTRACT FROM AUTHOR]

The paper illustrates a differential evolution (DE) algorithm to calculate the level-cuts of the fuzzy extension of a multidimensional real valued function to fuzzy numbers. The method decomposes the fuzzy extension engine into a set of "nested" min and max box-constrained optimization problems and uses a form of the DE algorithm, based on multi populations which cooperate during the search phase and specialize, a part of the populations to find the the global min (corresponding to lower branch of the fuzzy extension) and a part of the populations to find the global max (corresponding to the upper branch), both gaining efficienty from the work done for a level-cut to the subsequent ones. A special version of the algorithm is designed to the case of differentiable functions, for which a representation of the fuzzy numbers is used to improve efficiency and quality of calculations. The included computational results indicate that the DE method is a promising tool as its computational complexity grows on average superlinearly (of degree less than 1.5) in the number of variables of the function to be extended. [ABSTRACT FROM AUTHOR]

This paper presents a new way to solve the inverse problem of electrocardiography in terms of heart model parameters. The developed event estimation and recognition method uses a unified neural network (UNN)-based optimization system to determine the most relevant heart model parameters. A UNN-based preliminary ECG analyzer system has been created to reduce the searching space of the optimization algorithm. The optimal model parameters were determined by a relation between objective function minimization and robustness of the solution. The final evaluation results, validated by physicians, were about 96% correct. Starting from the fact that input ECGs contained various malfunction cases, such as Wolff-Parkinson-White (WPW) syndrome, atrial and ventricular fibrillation, these results suggest this approach provides a robust inverse solution, circumventing most of the difficulties of the ECG inverse problem. [ABSTRACT FROM AUTHOR]

This paper presents a Nonlinear Neuro-Fuzzy Network (NNFN) for equalization of channel distortion. The NFNN is constructed by using fuzzy rules that incorporate nonlinear functions. The structure and learning algorithm of NNFN are presented and the development of adaptive equalizer has been performed. The equalizer is applied for equalization of channel distortion of time-invariant and time-varying channels. The performance of NNFN based equalizer has been compared with the performance of other nonlinear equalizers. The obtained simulation results of NNFN based system satisfies the effectiveness of the proposed system in equalization of channel distortion. [ABSTRACT FROM AUTHOR]

In this paper a fuzzy system for detection of the triangular and rectangular traffic signs is presented. In many sign recognition systems reliable and fast shape detection is a prerequisite for successful classification. The proposed method operates on colour images in which it detects the characteristic points of signs by sets of fuzzy rules. These points are used then for extraction of the shapes that fulfil the fuzzy verification rules. The method allows very accurate and real-time detection of the planar triangles, inverted triangles, rectangles, and diamond shapes. The presented detector is a part of a driver-assisting-system for recognition of the road signs. The experimental results verify the method accuracy and robustness. [ABSTRACT FROM AUTHOR]

Neural networks have been used in the development of intelligent recognition systems that simulate our ability recognize patterns. However, rotated objects may cause incorrect identification by recognition systems. Our quick glance provides an overall approximation of a pattern regardless of noise or rotations. This paper proposes that the overall approximation of a pattern can be achieved via pattern averaging prior to training a neural network to recognize that pattern in various rotations. Pattern averaging provides the neural network with "fuzzy" rather than "crisp" representations of the rotated objects, thus, minimizing computational costs and providing the neural network with meaningful learning of various rotations of an object. The proposed method will be used to recognize rotated coins and is implemented to solve an existing problem where slot machines in Europe accept the new Turkish 1 Lira coin as a 2 Euro coin. [ABSTRACT FROM AUTHOR]

We describe in this paper the application of several neural network architectures to the problem of simulating and predicting the dynamic behavior of complex economic time series. We use several neural network models and training algorithms to compare the results and decide at the end, which one is best for this application. We also compare the simulation results with the traditional approach of using a statistical model. In this case, we use real time series of prices of consumer goods to test our models. Real prices of tomato and green onion in the U.S. show complex fluctuations in time and are very complicated to predict with traditional statistical approaches. [ABSTRACT FROM AUTHOR]

In this paper, we have presented new multivariate fuzzy time series (FTS) forecasting method. This method assume m-factors with one main factor of interest. Stochastic fuzzy dependence of order k is assumed to define general method of multivariate FTS forecasting and control. This new method is applied for forecasting total number of car road accidents causalities in Belgium using four secondary factors. Practically, in most of the situations, actuaries are interested in analysis of the patterns of casualties in road accidents. Such type of analysis supports in deciding approximate risk classification and forecasting for each area of a city. This directly affects the underwriting process and adjustment of insurance premium, based on risk intensity for each area. Thus, this work provides support in deciding the appropriate risk associated with an insured in a particular area. Finally, comparison is also made with most recent available work on fuzzy time series forecasting. [ABSTRACT FROM AUTHOR]

We describe in this paper the application of several neural network architectures to the problem of simulating and predicting the dynamic behavior of complex economic time series. We use several neural network models and training algorithms to compare the results and decide at the end, which one is best for this application. We also compare the simulation results with the traditional approach of using a statistical model. In this case, we use real time series of prices of consumer goods to test our models. Real prices of tomato and green onion in the U.S. show complex fluctuations in time and are very complicated to predict with traditional statistical approaches. [ABSTRACT FROM AUTHOR]

In this paper, the authors discuss a new synthesis approach to train associative memories, based on recurrent neural networks (RNNs). They propose to use soft margin training for associative memories, which is efficient when training patterns are not all linearly separable. On the basis of the soft margin algorithm used to train support vector machines (SVMs), the new algorithm is developed in order to improve the obtained results via optimal training algorithm also innovated by the authors, which are not fully satisfactory due to that some times the training patterns are not all linearly separable. This new algorithm is used for the synthesis of an associative memory considering the design based on a RNN with the connection matrix having upper bounds on the diagonal elements to reduce the total number of spurious memory. The scheme is evaluated via a full scale simulator to diagnose the main faults occurred in fossil electric power plants. [ABSTRACT FROM AUTHOR]

This paper discusses a decision support system based on fuzzy set techniques and tree search for movement planning in freight rail networks. The aim is to provide a decision making toot to help train dispatchers to plan and control rail traffic in real time. The system uses fuzzy rule-based decision procedures to drive train circulation as closed as possible to reference trajectories. A tree search algorithm is used to improve solution quality. A major characteristic of the system concerns its ability to provide feasible movement plans in real time. Incorporation of practical knowledge and real time response are essential requirements in real world freight railroads traffic management and control. [ABSTRACT FROM AUTHOR]

Fuzzy logic has been successfully used for nonlinear control systems. However, when the plant is complex or expert knowledge is not available, it is difficult to construct the rule bases of fuzzy systems. In this paper, we propose a new method of how to construct automatically the rule bases using fuzzy neural network. Whereas the conventional methods need the training data representing input-output relationship, the proposed algorithm utilizes the gradient of the performance index for the construction of fuzzy rules and the tuning of membership functions. Experimental results with the inverted pendulum show the superiority of the proposed method in comparison to the conventional fuzzy controller. [ABSTRACT FROM AUTHOR]

This paper presents a robust speed controller design method based on fuzzy logic control(FLC) for robust torsional vibration suppression control scheme in rolling mill drive system. This method proposes a torsional vibration suppression controller that comprises a reduced-order state feedback controller and a PI controller whose motor speed and observed torsional torque are fed back. By using the mechanical parameters estimated by an off-line recursive least square algorithm, a speed controller for torsional vibration suppression and its gains can be determined by FLC with the Kharitonov's robust control theory. This method can yield a robust stability with a specified stability margin and damping limit. Even if the parameters are varied within some specified range, the proposed control method guarantees a highly efficient vibration suppression. By using a fully digitalized 5.5 kW rolling mill drive system, the effectiveness and usefulness of the proposed scheme are verified and obtained experimental results. [ABSTRACT FROM AUTHOR]

Image segmentation is a classic problem in computer image comprehension and related fields. Up to now, there are not any general and valid partition methods which could satisfy different purposes, especially for medical images such as magnetic resonance images, which often corrupted by multiple imaging artifacts, for example intensity inhomogeneity, noise and partial volume effects. In this paper, we propose an improved fuzzy c-means image segmentation algorithm with more accurate results and faster computation. Considering two voxels with the same intensity belonging to the same tissue, we use q intensity levels instead of n intensity values in the objective function of the fuzzy c-means algorithm, which makes the algorithm clusters much faster since q is much smaller than n. Furthermore, a gain field is incorporate in the objective function to compensate for the inhomogeneity. In addition, we use c-means clustering algorithm to initialize the centroids. This can further accelerate the clustering. The test results show that the proposed algorithm not only gives more accurate results but also makes the computation faster. [ABSTRACT FROM AUTHOR]

This paper is concerned with a comparative study of the performance of fuzzy clustering algorithms Fuzzy C-Means (FCM), Gustafson-Kessel FCM (GK-FCM) and two variations of kernel-based FCM. One kernel-based FCM (KFCM) retains prototypes in the input space while the other (MKFCM) implicitly retains prototypes in the feature space. The two performance criteria used in the evaluation of the clustering algorithm deal with produced classification rate and reconstruction error. We experimentally demonstrate that the kernel-based FCM algorithms do not produce significant improvement over standard FCM for most data sets under investigation It is shown that the kernel-based FCM algorithms appear to be highly sensitive to the selection of the values of the kernel parameters. [ABSTRACT FROM AUTHOR]

Automated brain MR image segmentation is a challenging problem and received significant attention lately. Several improvements have been made to the standard fuzzy c-means (FCM) algorithm, in order to reduce its sensitivity to Gaussian, impulse, and intensity non-uniformity noises. In this paper we present a modified FCM algorithm, which aims accurate segmentation in case of mixed noises, and performs at a high processing speed. The proposed method extracts a scalar feature value from the neighborhood of each pixel, using a filtering technique that deals with both spatial and gray level distances. These features are classified afterwards using the histogram-based approach of the enhanced FCM classifier. The experiments using synthetic phantoms and real MR images show, that the proposed method provides better results compared to other reported FCM-based techniques. [ABSTRACT FROM AUTHOR]

Support vector machines (SVMs) are known to be useful for separating data into two classes. However, for the multiclass case where pairwise SVMs are incorporated, unclassifiable regions can exist. To solve this problem, Fuzzy support vector machines (FSVMs) was proposed, where membership values are assigned according to the distance between patterns and the hyperplanes obtained by the "crisp" SVM. However, they still may not give proper decision boundaries for arbitrary distributed data sets. In this paper, a density based fuzzy support vector machine (DFSVM) is proposed, which incorporates the data distribution in addition to using the memberships in FSVM. As a result, our proposed algorithm may give more appropriate decision boundaries than FSVM. To validate our proposed algorithm, we show experimental results for several data sets. [ABSTRACT FROM AUTHOR]