Search

This paper presents the advantage that knowledge-intensive activities, such as Scientific Conference Management, can take by the exploitation of expert components in the key tasks. Typically, in this domain the task of scheduling the activities and resources or the assignment of reviewers to papers is performed manually leading therefore to time-consuming procedures with high degree of inconsistency due to many parameters and constraints to be considered. The proposed systems, evaluated on real conference datasets, show good results compared to manual scheduling and assignment, in terms of both accuracy and reduction of runtime. [ABSTRACT FROM AUTHOR]

The aim of this paper is to give theoretical and experimental tools for measuring the driving force in evolving complex networks. First a discrete-time stochastic model framework is introduced to state the question of how the dynamics of these networks depend on the properties of the parts of the system. Then a method is presented to determine this dependence in the possession of the required data about the system. This measurement method is applied to the citation network of high energy physics papers to extract the in-degree and age dependence of the dynamics. It is shown that the method yields close to "optimal" results. [ABSTRACT FROM AUTHOR]

Nowadays amount of information on Internet is dramatically increasing. The ability of facilitating users to achieve useful information is more and more important for search field. CDSE, a model for the domain-based intelligent search engine is proposed in this paper. The model can help users to retrieve what they need by combining text classification with keywords extraction. Several algorithms that use key technologies are proposed, such as statistics, data mining and agents. Then a new criterion named ranking error is contributed to solve the problem of evaluation ranking inefficiency in traditional performance evaluation methodologies. The experimental results indicate that the proposed model can effectively improve retrieval precision and solve the problem of relevant document ranking behind in current search engine. [ABSTRACT FROM AUTHOR]

Reading and literature exploration are important tasks of scientific research. However, conventional retrieval systems provide limited support for these tasks by concentrating on identifying relevant materials. New generation systems should provide additional support functionality by focusing on analyzing and organizing the retrieved materials. A framework of literature exploration support systems is proposed. Techniques of granular computing are used to construct granular knowledge structures from the contents, structures, and usages of scientific documents. The granular knowledge structures provide a high level understanding of scientific literature and hints regarding what has been done and what needs to be done. As a demonstration, we examine granular knowledge structures obtained from an analysis of papers from two rough sets related conferences. [ABSTRACT FROM AUTHOR]

TSP(Traveling Salesman Problem) used widely for solving the optimization is the problem to find out the shortest distance out of possible courses where one starts a certain city, visits every city among N cities and turns back to a staring city. At this time, the condition is to visit N cities exactly only once. TSP is defined easily, but as the number of visiting cities increases, the calculation rate increases geometrically. This is why TSP is classified into NP-Hard Problem. Genetic Algorithm is used representatively to solve the TSP. Various operators have been developed and studied until now for solving the TSP more effectively. This paper applied the new Population Initialization Method (using the Random Initialization method and Induced Initialization method simultaneously), solved TSP more effectively, and proved the improvement of capability by comparing this new method with existing methods. Keywords: Genetic Algorithm, GA, Optimization, Initialization. [ABSTRACT FROM AUTHOR]

A method based on multi-feature and multi-stage verification is proposed in this paper. The direction and texture features obtained by Gabor filter are used for the first-stage verification. Then the sham dynamic features representing the writing force are used for the second-stage verification. Considering the number of samples for training is limited, the support vector machine (SVM) is selected as the classifier. The experimental results show that this method can not only increase the accuracy, but also satisfy the requirement of speed in real-time system as well. [ABSTRACT FROM AUTHOR]

Accurate cancer classification is very important to cancer diagnosis and treatment. As molecular information is increasing for the cancer classification, a lot of techniques have been proposed and utilized to classify and predict the cancers from gene expression profiles. In this paper, we propose a method based on speciated evolution for the cancer classification. The optimal combination among several feature-classifier pairs from the various features and classifiers is evolutionarily searched using the deterministic crowding genetic algorithm. Experimental results demonstrate that the proposed method is more effective than the standard genetic algorithm and the fitness sharing genetic algorithm as well as the best single classifier to search the optimal ensembles for the cancer classification. [ABSTRACT FROM AUTHOR]

This paper proposes a novel algorithm for nonnegative independent component analysis, which is based on minimizing the mutual information of the separated signals, and is truly insensitive to the particular underlying distribution of the source data. The unmixing system culminates to a novel neural network model. Compared with other algorithms for nonnegative ICA, the method proposed in this paper can work efficiently even in the case that the source signals are not well grounded, and that pre-whiting process is not needed. Finally, the experiments were performed on both simulating signals and mixtures of image data, the results indicate that the algorithm is efficient and effective. [ABSTRACT FROM AUTHOR]

This paper extends a survey on Networked Control Systems (NCSs) published in previous papers. General frameworks and some fundamental issues for the study of NCSs are proposed. A significant emphasis has been on developing control methodologies to study the character of NCSs. This survey paper presents recent NCSs control methodologies, some promising research directions are also covered. [ABSTRACT FROM AUTHOR]

Process models play important role in computer aided pro- cess engineering. Although the structure of these models are a priori known, model parameters should be estimated based on experiments. The accuracy of the estimated parameters largely depends on the information content of the experimental data presented to the parameter identification algorithm. Optimal experiment design (OED) can maximize the confidence on the model parameters. The paper proposes a new additive sequential evolutionary experiment design approach to maximize the amount of information content of experiments. The main idea is to use the identified models to design new experiments to gradually improve the model accuracy while keeping the collected information from previous experiments. This scheme requires an effective optimization algorithm, hence the main contribution of the paper is the incorporation of Evolutionary Strategy (ES) into a new iterative scheme of optimal experiment design (AS-OED). This paper illustrates the applicability of AS-OED for the design of feeding profile for a fed-batch biochemical reactor. [ABSTRACT FROM AUTHOR]

Traditional indexing methods face the difficulty of"curse of dimensionality" at high dimensionality. In this paper, the traditional vector approximation is improved. Firstly, it decreases the dimensions by LLE (locally-linear-embedding). As a result of it, a set of absolute low dimensions are gotten. Then, this paper uses the Gaussian mixture distribution and estimates the distribution through EM (expectation-maximization) method. The original data vectors are replaced by vector approximation. This approach gains higher efficiency and less run time. The experiments show a remarkable reduction of I/O. They also show an improvement on the indexing performance and then speed the image retrieval. [ABSTRACT FROM AUTHOR]

This paper presents the integration of linguistic knowledge in learning semantic user profiles able to represent user interests in a more effective way with respect to classical keyword-based profilesThis research was partially funded by the EC under the 6th FP IST IP VIKEF No. 507173, Priority 2.3.1.7 Semantic Based Knowledge Systems (www.vikef.net) and under the DELOS NoE on Digital Libraries, Priority: Technology-enhanced Learning and Access to Cultural Heritage - Contract n. G038-507618 (www.delos.info). The work was supported also by the research project "Modelli avanzati di personalizzazione per biblioteche elettroniche", year 2005, funded by the University of Bari.. Semantic profiles are obtained by integrating a naïve Bayes approach for text categorization with a word sense disambiguation strategy based on the WordNet lexical database (Section 2). Semantic profiles are exploited by the "conference participant advisor" service in order to suggest papers to be read and talks to be attended by a conference participant. Experiments on a real dataset show the effectiveness of the service (Section 3). [ABSTRACT FROM AUTHOR]

The inherent nonlinear of switched reluctance motor (SRM) makes it hard to get a good performance by using the conventional PID controller to the speed control of SRM. This paper develops a radial basis function (RBF) artificial neural network (ANN) nonlinear prediction model based adaptive PID controller for SRM. ANN, under certain condition, can approximate any nonlinear function with arbitrary precision. It also has a strong ability of adaptive, self-learning and self-organization. So, combining it with the conventional PID controller, a neural network based adaptive PID controller can be developed. Appling it to the speed control of SRM, a good control performance can be gotten. At the same time, the nonlinear mapping property and high parallel operation ability of ANN make it suitable to be applied to establish nonlinear prediction model performing parameter prediction. In this paper, two ANN - NNC and NNI are employed. The former is a back propagation (BP) ANN with sigmoid activation function. The later is an ANN using RBF as activation function. The former is used to adaptively adjust the parameters of the PID controller on line. The later is used to establish nonlinear prediction model performing parameter prediction. Compared with BP ANN with sigmoid activation function, the RBF ANN has a more fast convergence speed and can avoid getting stuck in a local optimum. Through parameter prediction, response speed of the system can be improved. To increase the convergence speed of ANN, an adaptive learning algorithm is adopted in this paper that is to adjust the learning rate according to the error. This can increase the convergence speed of ANN and make the system response quick. The experimental results demonstrate that a high control performance is achieved. The system responds quickly with little overshoot. The steady state error is zero. The system shows robust performance to the load torque disturbance. [ABSTRACT FROM AUTHOR]

The use of an ontology in order to provide a mechanism to enable machine reasoning has continuously increased during the last few years. This paper proposed an automated method for document classification using an ontology, which expresses terminology information and vocabulary contained in Web documents by way of a hierarchical structure. Ontology-based document classification involves determining document features that represent the Web documents most accurately, and classifying them into the most appropriate categories after analyzing their contents by using at least two pre-defined categories per given document features. In this paper, Web documents are classified in real time not with experimental data or a learning process, but by similarity calculations between the terminology information extracted from Web documents and ontology categories. This results in a more accurate document classification since the meanings and relationships unique to each document are determined. [ABSTRACT FROM AUTHOR]

The research work in this paper focuses on solving the critical problems in Web Service discovery such as how to locate functionality-desired Web Services and how to select the best one from large numbers of functionality-similar Web Services. The semantic description and quality description of Web Service based on ontology proposed in this paper, provides a consistent description of different kinds of Web Service, which is used as the basis of service matching. Then the matching model of Web Service is discussed in detail, which consists of semantic similarity matching according to functionality of Web Service, and semantic filtering according to quality of Web Service. Using the descriptions and matching approach, the performance of Web Service discovery can be improved by increasing the precision and recall of Web Service searching. [ABSTRACT FROM AUTHOR]

This paper illustrates an algorithm for osteosarcoma segmentation, using vectorial fuzzy-connectedness segmentation, and coming up with a methodology which can be used to segment some distinct tissues of osteosarcoma such as tumor, necrosis and parosteal sarcoma from 3D vectorial images. However, fuzzy-connectedness segmentation can be successfully used only in connected regions. In this paper, some improvements have been made to segment the interested tissues which are distributed in disconnected regions. And the paper speeds up the process of segmentation by segmenting two osteosarcoma tissues simultaneously. The methology has been applied to a medical image analysis system of osteosarcoma segmentation and 3D reconstruction, which has been put into practical use in some hospitals. [ABSTRACT FROM AUTHOR]

This paper suggests an automated method for document classification using an ontology, which expresses terminology information and vocabulary contained in Web documents by way of a hierarchical structure. Ontologybased document classification involves determining document features that represent the Web documents most accurately, and classifying them into the most appropriate categories after analyzing their contents by using at least two pre-defined categories per given document features. In this paper, Web documents are classified in real time not with experimental data or a learning process, but by similar calculations between the terminology information extracted from Web texts and ontology categories. This results in a more accurate document classification since the meanings and relationships unique to each document are determined. Keywords: Document classification, Ontology, Web Page classification. [ABSTRACT FROM AUTHOR]

This chapter contains papers presented in the workshop on Text Mining and Applications (TeMA 2005), organized in the framework of the Portuguese Association for Artificial Intelligence conference (EPIA). This workshop is aimed at attracting quality papers and enhancing the knowledge in this area. 27 papers were submitted. From these, 9 papers were selected for publication in this Springer volume. These numbers show current importance of this field in AI and suggest that the organization of equivalent events in future EPIA editions should be pursued. First paper works on bilingual lexical acquisition from non parallel corpora, applicable in Machine Translation. Second paper describes work on Text Summarization. Third uses Transformation Based learning for NP Identification applied to Portuguese.Fourth uses linguistic knowledge for passage retrieval and question answering. Fifth describes the use of weakly supervised learning for extraction of semantic patterns. Sixth paper proposes a method for semantic indexing and evaluates it on traditional Information Retrieval tasks. Seventh works on unsupervised language independent extraction of multi-word terms, applicable in multiple domains and evaluates their results for Slovene and English. Eighth paper presents a variant of a known method for Anaphora resolution, adapted to Portuguese. Last paper proposes a stemmer for Brazilian Portuguese. [ABSTRACT FROM AUTHOR]

A Personal Assistant (PA) agent is a software agent capable of helping people to handle tasks in their workplace. The paper proposes a memory mechanism for personal assistant agents in order to enhance agent intelligence while working with the user or with other agents. Inspired by a case memory model in the domain of Case-Based Reasoning (CBR), this paper endows PA agents with a case memory mechanism, which results in improved PA agents: MemoPAs. We present the memory mechanism of MemoPA in detail, and report a first implementation of the method. Finally, future work is outlined for improving the memory mechanism. [ABSTRACT FROM AUTHOR]

A top-down experimental procedure for defect type recognition of epoxy-resin transformers by Partial Discharge (PD) is proposed. Most of the PD detection methods could be performed only at the shutdown period of equipments. By using Acoustic Emission (AE), the real-time and online detection could be reachable. Therefore, this paper conducted high voltage test of pre-faulty transformers and measured those PD signals for recognition needed. Afterward, the selected features that proposed in this paper can be extracted from these collected PD signals. According to these features, effective identification of their faulty types can be done using the proposed particle swarm optimization combined with neural network. Finally, with a view to apply in field, this research adds different noise levels to distort the original data. These distorted data are entered for subsequent testing. Research shows encouraging results that with 30% noise per discharge count, an 80% successful recognition rate can be achieved. [ABSTRACT FROM AUTHOR]

Electromagnetism-like algorithm (EM) is a population-based meta-heuristic which has been proposed to solve continuous problems effectively. In this paper, we present a new meta-heuristic that uses the EM methodology to solve the single machine scheduling problem. Single machine scheduling is a combinatorial optimization problem. Schedule representation for our problem is based on random keys. Because there is little research in solving the combinatorial optimization problem (COP) by EM, the paper attempts to employ the random-key concept enabling EM to solve COP in single machine scheduling problem. We present a hybrid algorithm that combines the EM methodology and genetic operators to obtain the best/optimal schedule for this single machine scheduling problem, which attempts to achieve convergence and diversity effect when they iteratively solve the problem. The objective in our problem is minimization of the sum of earliness and tardiness. This hybrid algorithm was tested on a set of standard test problems available in the literature. The computational results show that this hybrid algorithm performs better than the standard genetic algorithm. [ABSTRACT FROM AUTHOR]

Upon close examination of a set of industrial data from a large scale power plant, time varying behavior are discovered. If a fixed model is applied to monitor this process, false alarms will be inevitable. This paper suggests the use of adaptive models to cope with such situation. A recently proposed technique, fast algorithm for Moving Window Principal Component Analysis (MWPCA) was employed because of its following strength: (i) the ability in adapting process changes, (ii) the conceptual simplicity, and (iii) its computational efficiency. Its advances in fault detection is demonstrated in the paper by comparing with the conventional PCA. In addition, this paper proposed to plot the scaled variables in conjunction with MWPCA for fault diagnosis, which is proved to be effective in this application. [ABSTRACT FROM AUTHOR]

PVM (Parallel virtual machine) library is a tool which used processes large amounts of data sets. This paper wants to achieve a high performance solution that exploits PVM library and parallel computers to solve ICA (Independent Component Analysis) problem. The paper presents parallel power ICA implementations to decomposition data sets. Power iteration (PI) is an algorithm for independent component analysis, which has some desired features. It has higher performance and data capacity than current sequential implementations. This paper, we show the power iteration algorithm which learning updating is in the form of matrix transformation . From power iteration algorithm, we develop parallel power iteration algorithm and implement parallel component decomposition solution. At last, experimental results, analysis and future plans are presented. [ABSTRACT FROM AUTHOR]

In this paper, global exponential stability of Cohen-Grossberg neural networks with reaction-diffusion and Dirichlet boundary conditions is considered by using an approach based on the delay differential inequality and the fixed-point theorem. Some sufficient conditions are obtained to guarantee that the reaction-diffusion Cohen-Grossberg neural networks are globally exponentially stable. The results presented in this paper are the improvement and extension of the existed ones in some existing works. [ABSTRACT FROM AUTHOR]

This paper analyzes global convergence and learning parameters of the back-propagation algorithm for quadratic functions. Some global convergence conditions of the steepest descent algorithm are obtained by directly analyzing the exact momentum equations for quadratic cost functions. In addition, in order to guarantee the convergence for a given learning task, the method is obtained to choose the proper learning parameters. The results presented in this paper are the improvement and extension of the existed ones in some existing works. [ABSTRACT FROM AUTHOR]

The automatic relevance determination (ARD) shows good performance in many applications. Recently, it has been applied to brain current estimation with the variational method. Although people who use the ARD tend to pay attention to one benefit of the ARD, sparsity, we, in this paper, focus on another benefit, generalization. In this paper, we clarify the generalization error of the ARD in the case that a class of prior distributions is used, and show that good generalization is caused by singularities of the ARD. Sparsity is not observed in that case, however, the mechanism that the singularities provide good generalization implies the mechanism that they also provide sparsity. [ABSTRACT FROM AUTHOR]

This paper presents iris recognition for personal identification using neural networks. Iris recognition system consists of localization of the iris region and generation of data set of iris images and then iris pattern recognition. One of the problems in iris recognition is fast and accurate localization of the iris image. In this paper, fast algorithm is used for the localization of the inner and outer boundaries of the iris region. Located iris is extracted from an eye image, and, after normalization and enhancement it is represented by a data set. Using this data set a neural network is applied for the classification of iris patterns. Results of simulations illustrate the effectiveness of the neural system in personal identification. [ABSTRACT FROM AUTHOR]

This article deals with the identification of hinging hyperplane models. This type of non-linear black-box models is relatively new, and its identification is not thoroughly examined and discussed so far. They can be an alternative to artificial neural nets but there is a clear need for an effective identification method. This paper presents a new identification technique for that purpose based on a fuzzy clustering technique called Fuzzy c-Regression Clustering. To use this clustering procedure for the identification of hinging hyperplanes there is a need to handle restrictions about the relative location of the hyperplanes: they should intersect each other in the operating regime covered by the data points. The proposed method recursively identifies a hinging hyperplane model that contains two linear submodels by partitioning of the operating region of one local linear model resuling in a binary regression tree. Hence, this paper proposes a new algorithm for the identification of tree structured piecewise linear models, where the branches correspond to linear division of the operating regime based on the intersection of two local linear models. The effectiveness of the proposed model is demonstrated by a dynamic model identification example. [ABSTRACT FROM AUTHOR]

The aim of this paper is to compare the effectiveness of various computational intelligence approaches applied to the task of retrieving musical rhythm from musical symbolic files. The study presented in this paper describes how Artificial Neural Networks and Rough Sets can be used for searching the metric structure of musical files. The described approaches are based on examining physical attributes of sound that are most significant in determining the placement of a particular sound in the accented location of a musical piece. The results of the experiments show that the approach based solely on duration is sufficient enough to retrieve the metric structure of rhythm from musical files. [ABSTRACT FROM AUTHOR]

This paper is an extension of our earlier work in combining strengths of rough set theory and neuro-fuzzy decision trees in classifying software defect data. The extension includes the application of a rough-fuzzy classification trees to classifying defects. We compare classification results for five methods: rough sets, neuro-fuzzy decision trees, partial decision trees, rough-neuro-fuzzy decision trees and rough-fuzzy classification trees. The analysis of the results include a paired t-test for accuracy and number of rules. The results demonstrate that there is improvement in classification accuracy with the rough fuzzy classification trees with a minimal set of rules. The contribution of this paper is a comparative study of several hybrid approaches in classifying software defect data. [ABSTRACT FROM AUTHOR]

Currently used data-mining algorithms treat data globally. Nevertheless, with such methods, potentially useful knowledge that relates to local phenomena may be undetected. In this paper, we introduce new patterns in a form of local frequent events and epochs, boundaries of which correspond to discovered changes in a data stream. A local frequent event is an event which occurs in some period of time frequently, but not necessarily in the whole data stream. Such an event will be called a frequent event in a data stream. An epoch is understood as a sufficiently large group of frequent events that occur in a similar part of the data stream. The epochs are defined in such a way that they do not overlap are separated by so called change periods. In the paper, we discuss some potential applications of the proposed knowledge. Preliminary experiments are described as well. [ABSTRACT FROM AUTHOR]

In T −fuzzy rough sets a fuzzy T −similarity relation is employed to describe the similar degree between two objects and to construct lower and upper approximations for arbitrary fuzzy sets. The existing researches on T −fuzzy rough sets mainly concentrate on constructive and axiomatic approaches of lower and upper approximation operators. In this paper we define attribute reduction based on T −fuzzy rough sets. The structure of proposed attribute reduction is investigated in detail by the approach of discernibility matrix. At last an example is proposed to illustrate our idea in this paper. [ABSTRACT FROM AUTHOR]

The subject-matter of the consideration touches the problem of vagueness. The notion of the rough set, originated by Zdzisław Pawlak, was constructed under the influence of vague information and methods of shaping systems of notions leading to conceptualization and representation of vague knowledge, so also systems of their scopes as some vague sets. This paper outlines some direction of searching for a solution to this problem. In the paper, in connection to the notion of the rough set, the notion of a vague set is introduced. Some operations on these sets and their properties are discussed. The considerations intend to take into account a classical approach to reasoning, based on vague premises, and suggest finding a logic of vague sentences as a non-classical logic in which all counterparts of tautologies of classical logic are laws. [ABSTRACT FROM AUTHOR]

We contribute two new algorithms in this paper called PImGA and PIrlEA respectively in which we construct populations online in each iteration. Every iteration process in these two algorithms does not like the normal EA and GA in which they employ the inefficient value iteration method in general, instead of, in this paper, we employ the efficient policy iteration as the computation method for searching optimal control actions or policies. Meanwhile,these algorithms also do not like general EA and GA for selection operator to get a optimal policy, instead of we make the Agent learning a good or elite policy from its parents population. The resulted policy will be as one of elements of the next population. Because this policy is obtained by taking optimal reinforcement learning algorithm and greedy policy, the new population always can be constructed by applying better policies than its parents, that is to say, the child or offspring will inherit parents' good or elite abilities. Intuitively, for a finite problem, the resulted population from simulation will accommodate the near optimal policies after a number of iterations. Our experiments show that the algorithms can work well. [ABSTRACT FROM AUTHOR]

The particle swarm optimization (PSO) is a stochastic evolutionary computation technique based on the behavior of swarms that can be used to optimize objects with complex search spaces. However, it has been observed that its performance varies duo to the dimensionality of the object and the location of the global optimum in the search space. This paper introduces a "random" velocity boundary condition to address the problem, where the velocity boundary alters randomly to prevent the velocity of a particle from stopping on a same boundary during the evolution. Simulation results on two benchmark functions with 30 and 300 dimensionalities and three types of locations of the global optimum solutions in the search spaces have shown that with the proposed "random" velocity boundary condition, a highly competitive optimization performance can be obtained for PSO regardless of the dimensionality and the location of the global optimum solution. [ABSTRACT FROM AUTHOR]

Many function-approaching methods such as neural network, fuzzy method are used in reinforcement learning methods for solving its huge problem space dimensions. This paper presents a novel ART2 neural network based reinforcement learning method (ART2-RL) to solve the space problem. Because of its adaptive resonance characteristic, ART2 neural network is used to process the space measurement of reinforcement learning and improve the learning speed. This paper also gives the reinforcement learning algorithm based on ART2. A simulation of path planning of mobile robot has been developed to prove the validity of ART2-RL. As the complexity of the simulation increased, the result shows that the number of collision between robot and obstacles is effectively decreased; the novel neural network model provides significant improvement in the space measurement of reinforcement learning. [ABSTRACT FROM AUTHOR]

In this paper, we present a novel image digital watermarking technique based on Kernel Independent Component Analysis (KICA). Use the nice characteristic of the KICA, which can results the blind separation of nonlinearly mixed signals, the imperceptibility and robustness requirements of watermarks are fulfilled and optimized. In the proposed scheme, the watermark image is first transformed by Arnold method, and then embedded into the lowest frequency subband in DWT domain. The recovery of owner's image is turning the watermarked image into DWT domains then use KICA to extract the watermark. Finally the watermark is transformed by Arnold method again, so we can get the original watermark image. Experimental results show that the proposed method gains better performance in robustness than that of ICA with respect to traditional image processing including cropping, filtering, add noise and JPEG image compression. [ABSTRACT FROM AUTHOR]

Knowledge discovery methods used to find relationships among software engineering data and the extraction of rules have gained increasing importance in recent years. These methods have become necessary for improvements in the quality of the software product and the process. The focus of this paper is a first attempt towards combining strengths of rough set theory and neuro-fuzzy decision trees in classifying software defect data. We compare classification results for four methods: rough sets, neuro-fuzzy decision trees, partial decision trees, rough-neuro-fuzzy decision trees. The analysis of the results include a family-wise 10 fold paired t-test for accuracy and number of rules. The contribution of this paper is the application of a hybrid rough-neuro-fuzzy decision tree method in classifying software defect data. [ABSTRACT FROM AUTHOR]

The paper proposes the use of the multilayer perceptron model to the problem of detecting ham and spam e-mail patterns. It also proposes an intensive use of data pre-processing and feature selection methods to simplify the task of the multilayer perceptron in classifying ham and spam e-mails. The multilayer perceptron is trained and assessed on patterns extracted from the SpamAssassin Public Corpus. It is required to classify novel types of ham and spam patterns. The results are presented and evaluated in the paper. [ABSTRACT FROM AUTHOR]

In singular models, the Bayes estimation, commonly, has the advantage of the generalization performance over the maximum likelihood estimation, however, its accurate approximation using Markov chain Monte Carlo methods requires huge computational costs. The variational Bayes (VB) approach, a tractable alternative, has recently shown good performance in the automatic relevance determination model (ARD), a kind of hierarchical Bayesian learning, in brain current estimation from magnetoencephalography (MEG) data, an ill-posed linear inverse problem. On the other hand, it has been proved that, in three-layer linear neural networks (LNNs), the VB approach is asymptotically equivalent to a positive-part James-Stein type shrinkage estimation. In this paper, noting the similarity between the ARD in a linear problem and an LNN, we analyze a simplified version of the VB approach in the ARD. We discuss its relation to the shrinkage estimation and how ill-posedness affects learning. We also propose the algorithm that requires simpler computation than, and will provide similar performance to, the VB approach. [ABSTRACT FROM AUTHOR]

This paper presents an entirely original concept of IT systems construction never published before. It supports the strategic business decision-making processes based on cognitive analysis methods. The essence of this new approach is in that the automatic understanding methods, tested previously in the area of medical image interpretation, enhancing the traditional set of automatic analysis and automatic classification methods, will be used to develop new generation business IT systems. This paper proposes a holistic concept of a new IT economy system, which is a significantly enhanced DSS-type system (Decision Support Systems). The essence of innovations introduced in this paper is in adding understandable business data to the analysis process of the said data. Such a system is essentially different from all known DSS-type systems, it is also different from all approaches to intelligent IT system construction, as described in the literature like, for example, those based on neural networks or in the use of expert systems. In this paper we therefore propose an acronym for the new system labelled UBMSS ( Understanding Based Managing Support Systems). Cognitive methods, on which the UBMSS concept and construct are based, copy the psychological and neurophysiological processes of understanding the analysed data, as they take place in the brain of a competent and particularly gifted man. [ABSTRACT FROM AUTHOR]

n this paper investigate how to aggregation method from face recognition varying environments. Face Images clustering is enhanced face recognition performance. Face image is clustered several cluster unsupervised or statistical method and we recognize using correlation between clusters. In this paper we adopted recognition algorithm by aggregation method. In this paper we present the recognition system using the table of fitness correlations between clusters for combining the results from the individual clusters. By training the different classifiers with different clusters of training data and adopting fusion method considering fitness correlation between clusters we found out better recognition performance than combining classifiers fed with same data. [ABSTRACT FROM AUTHOR]

Gene libraries are a biological mechanism for generating combinatorial diversity in the immune system. However, they also bias the antibody creation process, so that they can be viewed as a way of guiding lifetime learning mechanisms. In this paper we examine the implications of this view, by examining coverage, avoidance of self, clustering and diversity. We show how gene libraries may improve both computational expense and performance, and present an analysis which suggests how they might do it. We suggest that gene libraries: provide combinatorial efficiency; improve coverage; reduce the cost of negative selection; and allow targeting of fixed antigen populations. Keywords: gene libraries, artificial immune systems, antibodies, diversity, Baldwin effect, lifetime learning. [ABSTRACT FROM AUTHOR]

In this paper, we originally propose a music recommendation agent in a smart office to recommend music for users. Personal tastes are diverse, even the same person may have different preferences in different situations. Hence, our music recommendation is not only based on the user favorite genres but also the current mood of users. By collecting and analyzing the contextual information of users, the agent can automatically senses the mood of users and recommends music. [ABSTRACT FROM AUTHOR]

In this paper, we propose face feature selection and classifier selection method for face image group according illuminant. In knowledge based, we stored context and weight for feature points and selected classifier for context. This context is distinguished the face images having varying illumination. This context knowledge can be accumulated and used later. Therefore we designed the face recognition system by using evolution method and efficient face feature point selection. It can improve its performance incrementally using proposed algorithm. And we proposed efficient context modeling method by using SOM. For context awareness, we made artificial face images from FERET fa dataset and divided several group. Therefore we improved face recognition ratio using adaptable classifier, feature and weight for feature points. [ABSTRACT FROM AUTHOR]

A novel contourlet transform based fusion algorithm for nighttime driving image is proposed in this paper. Because of advantages of the contourlet transform in dealing with the two or higher dimensions singularity or the image salient features, such as line, curve, edge and etc., each of the accurately registered images is decomposed into a low frequency subband image and a sets of high frequency subband images with various multiscale, multidirectional local salient features. By using different fusion rules for the low frequency subband image and high frequency subband images, respectively, the fused coefficients are obtained. Then, the fused image is generated by the inverse contourlet transform. The simulation results indicate that the proposed method outperforms the traditional wavelet packet transform based image fusion method. [ABSTRACT FROM AUTHOR]

One of the most crucial and complex steps of object-oriented system design lies in the transition needed to go from system behavior (defined by means of scenario models) to component behavior (described by means of communicating hierarchical state machine models).This paper presents a re-verse approach for generating hierarchical UML statechart diagrams. Firstly, we put forward a generation algorithm for a flat statechart diagram based on the BK-algorithm, which is validated useful in our reverse engineering tool XDRE by generating UML statechart diagrams from a set of UML sequence diagrams. Secondly, according to UML composite state, an automatic approach of introducing hierarchy to the generated flat statechart diagrams is proposed and implemented. Finally, systematic experiment is conducted in the paper in order to verify the validity of this approach. [ABSTRACT FROM AUTHOR]

Rough set theory has been proposed by Pawlak as a tool for dealing with the vagueness and granularity in information systems. The core concepts of classical rough sets are lower and upper approximations based on equivalence relations. This paper studies arbitrary binary relation based generalized rough sets. In this setting, a binary relation can generate a lower approximation operation and an upper approximation operation. We prove that such a binary relation is unique, since two different binary relations will generate two different lower approximation operations and two different upper approximation operations. This paper also explores the relationships between the lower or upper approximation operation generated by the intersection of two binary relations and those generated by these two binary relations, respectively. Keyword: Rough set, Lower approximation, Upper approximation, Binary relation, Fuzzy set, Granular computing. [ABSTRACT FROM AUTHOR]

Although many works have been done in recent years for the designing Fuzzy-Neural Networks (FNN) from input-output data, the results concerning how to analyze the performance of some methods from a rigorous mathematical point of view are somewhat few. In this paper, the approximation bound for the Table Look-up Scheme with the Bell Membership Function is established. The detailed formulas of the error bound between the nonlinear function to be approximated and the FNN system designed based on the input-output data are derived. [ABSTRACT FROM AUTHOR]

The accurate rotor position information is very important for high performance operation of switched reluctance motor (SRM). Traditionally, there is a mechanical rotor position sensor. But this reduces the reliability and increases cost and size of SRM. In order to overcome the disadvantage of mechanical rotor position sensor, a sensorless operation method of SRM based on adaptive network based fuzzy inference system (ANFIS) is developed in this paper. The rotor position can be estimated by using the unique relationship between rotor position, flux linkage and phase current. In this paper, the ANFIS is used to map this relationship. Among the sensorless position estimation method, approach based on fuzzy neural network (FNN) is one of the promising methods. By combining the benefits of artificial neural network (ANN) and fuzzy inference system (FIS) in a single model, the ANFIS shows characteristics of fast and accurate learning, the ability of using both linguistic information and data information and good generalization capability. For its antecedents are fuzzy sets, the noise in the input signals can be restrained. This approach shows a characteristic of robustness. For its consequent is in linear function of input variables, it has a simple structure and low computation complexity. So, it is well suited to be applied on-line. Applying it to the rotor position estimation, a high accuracy and robust sensorless rotor position estimator is presented. The experimental results proved the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]