Search

When different subsamples of the same data set are used to induce classification trees, the structure of the built classifiers is very different. The stability of the structure of the tree is of capital importance in many domains, such as illness diagnosis, fraud detection in different fields, customer's behaviour analysis (marketing), etc, where comprehensibility of the classifier is necessary. We have developed a methodology for building classification trees from multiple samples where the final classifier is a single decision tree (Consolidated Trees). The paper presents an analysis of the structural stability of our algorithm versus C4.5 algorithm. The classification trees generated with our algorithm, achieve smaller error rates and structurally more steady trees than C4.5 when using resampling techniques. The main focus on this paper is showing how Consolidated Trees built with different sets of subsamples tend to converge to the same tree when the number of used subsamples is increased. [ABSTRACT FROM AUTHOR]

We study scheduling problems in battery-operated computing devices, aiming at schedules with low total energy consumption. While most of the previous work has focused on finding feasible schedules in deadline-based settings, in this paper we are interested in schedules that guarantee good response times. More specifically, our goal is to schedule a sequence of jobs on a variable speed processor so as to minimize the total cost consisting of the power consumption and the total flow time of all the jobs. We first show that when the amount of work, for any job, may take an arbitrary value, then no online algorithm can achieve a constant competitive ratio. Therefore, most of the paper is concerned with unit-size jobs. We devise a deterministic constant competitive online algorithm and show that the offline problem can be solved in polynomial time. [ABSTRACT FROM AUTHOR]

We study route selection for packet switching in the competitive throughput model. In contrast to previous papers which considered competitive algorithms for packet scheduling, we consider the packet routing problem (output port selection in a node). We model the node routing problem as follows: a node has an arbitrary number of input ports and an arbitrary number of output queues. At each time unit, an arbitrary number of new packets may arrive, each packet is associated with a subset of the output ports (which correspond to the next edges on the allowed paths for the packet). Each output queue transmits packets in some arbitrary manner. Arrival and transmission are arbitrary and controlled by an adversary. The node routing algorithm has to route each packet to one of the allowed output ports, without exceeding the size of the queues. The goal is to maximize the number of the transmitted packets. In this paper, we show that all non-refusal algorithms are 2-competitive. Our main result is an almost optimal -competitive algorithm, for a large enough queue size. For packets with arbitrary values (allowing preemption) we present a 2-competitive algorithm for any queue size. [ABSTRACT FROM AUTHOR]

The common theoretical model adopted in recent studies on algorithms for systems of autonomous mobile robots assumes that the positional input of the robots is obtained by perfectly accurate visual sensors, that robot movements are accurate, and that internal calculations performed by the robots on (real) coordinates are perfectly accurate as well. The current paper concentrates on the effect of weakening this rather strong set of assumptions, and replacing it with the more realistic assumption that the robot sensors, movement and internal calculations may have slight inaccuracies. Specifically, the paper concentrates on the ability of robot systems with inaccurate sensors, movements and calculations to carry out the task of convergence. The paper presents several impossibility results, limiting the inaccuracy allowing convergence. The main positive result is an algorithm for convergence under bounded measurement, movement and calculation errors. [ABSTRACT FROM AUTHOR]

This paper describes circuit evolutionary experiments at extreme low temperatures, including the test of all system components at this extreme environment (EE). In addition to hardening-by-process and hardening-bydesign, "hardening-by-reconfiguration", when applicable, could be used to mitigate drifts, degradation, or damage on electronic devices (chips) in EE, by using re-configurable devices and an adaptive self-reconfiguration of their circuit topology. Conventional circuit design exploits device characteristics within a certain temperature/radiation range; when that is exceeded, the circuit function degrades. On a reconfigurable device, although component parameters change in EE, a new circuit design, suitable for new parameter values, may be mapped into the reconfigurable structure to recover the initial circuit function. This paper demonstrates this technique for circuit evolution and recovery at liquid nitrogen temperatures (-196.6∘C). In addition, preliminary tests are performed to assess the survivability of the evolutionary processor at extreme low temperatures. [ABSTRACT FROM AUTHOR]

The ideas presented in this paper are results of a new research project, "Reconfigurable POEtic Tissue". The goal of the project was the development of a hardware platform capable of implementing bio-inspired systems, in digital hardware. In particular, the final hardware device, while similar to other FPGAs, was designed with a number of novel features which facilitate faulttolerance. These include dynamic reconfiguration and on-chip re-programming. This paper considers these features in the context of fault-tolerant system design and shows how an ensemble of different, but often complementary, techniques might be produced using these novel device features. Such characteristics are crucial for many control systems, particularly with safety implications. [ABSTRACT FROM AUTHOR]

Hardware immune systems have been studied with some initial achievements in recent years. Hardware immune systems are inspired by biological immune systems and they are expected to have many interesting characteristics, such as self-adaptive, self-learning and fault tolerant abilities. However, as novel intelligent systems, hardware immune systems are faced with many problems. This paper focuses on autoimmunization that is an inevitable problem when designing a complex hardware immune system. After the costimulation mechanism of biological immune system is simply introduced as a metaphor, a novel self-adaptive and self-tolerant algorithm for hardware immune systems is proposed in this paper. Inspired by the co-stimulation mechanism, the algorithm endows hardware immune systems with the capability of self-tolerance by automatically updating detector set and making the self set more complete. It can increase the accuracy of detection and decrease the rate of false positive effectively. Results of simulation experiments demonstrate the validity of this algorithm. [ABSTRACT FROM AUTHOR]

The self-assembly process for bottom-up construction of nanostructures is of key importance to the emerging scientific discipline Nanoscience. However, self-assembly at the molecular scale is prone to a quite high rate of error. Such high error rate is a major barrier to large-scale experimental implementation of DNA tiling. The goals of this paper are to develop theoretical methods for compact error-resilient self-assembly and to analyze these methods by stochastic analysis and computer simulation. Prior work by Winfree provided an innovative approach to decrease tiling self-assembly errors without decreasing the intrinsic error rate ε of assembling a single tile. However, his technique resulted in a final structure that is four times the size of the original one. This paper describes various compact error-resilient tiling methods that do not increase the size of the tiling assembly. These methods apply to assembly of boolean arrays which perform input sensitive computations (among other computations). Our 2-way (3-way) overlay redundancy construction drops the error rate from ε to approximately ε2 (ε3), without increasing the size of the assembly. These results were further validated using stochastic analysis and computer simulation. [ABSTRACT FROM AUTHOR]

$p\ {\overset{x}{\rightarrow}}\ q$ This paper discusses a problem of checking the structure freeness of DNA and RNA sequences. A set R of sequences over Σ is said to be structure free if for every α ∈ R, the minimum free energy of α is greater than or equal to 0. It was open whether or not there exists a polynomial time algorithm for testing the structure freeness of a given regular set R. In this paper, we will solve this problem. Keywords: Molecular Computing, DNA Computing, Sequence Design, Secondary Structures. [ABSTRACT FROM AUTHOR]

Conflicts naturally arise in optimistically replicated systems. The common way to detect update conflicts is via version vectors, whose storage and communication overhead are number of replicas × number of objects. These costs may be prohibitive for large systems. This paper presents predecessor vectors with exceptions (PVEs), a novel optimistic replication technique developed for Microsoft's WinFS system. The paper contains a systematic study of PVE's performance gains over traditional schemes. The results demonstrate a dramatic reduction of storage and communication overhead in normal scenarios, during which communication disruptions are infrequent. Moreover, they identify a cross-over threshold in communication failure-rate, beyond which PVEs loses efficiency compared with traditional schemes. [ABSTRACT FROM AUTHOR]

This paper studies implementations of concurrent objects that exploit the absence of step contention. These implementations use only reads and writes when a process is running solo. The other processes might be busy with other objects, swapped-out, failed, or simply delayed by a contention manager. We study in this paper two classes of such implementations, according to how they handle the case of step contention. The first kind, called obstruction-free implementations, are not required to terminate in that case. The second kind, called solo-fast implementations, terminate using powerful operations (e.g., C&S). We present a generic obstruction-free object implementation that has a linear contention-free step complexity (number of reads and writes taken by a process running solo) and uses a linear number of read/write objects. We show that these complexities are asymptotically optimal, and hence generic obstruction-free implementations are inherently slow. We also prove that obstruction-free implementations cannot be gracefully degrading, namely, be nonblocking when the contention manager operates correctly, and remain (at least) obstruction-free when the contention manager misbehaves. Finally, we show that any object has a solo-fast implementation, based on a solo-fast implementation of consensus. The implementation has linear contention-free step complexity, and we conjecture solo-fast implementations must have non-constant step complexity, i.e., they are also inherently slow. [ABSTRACT FROM AUTHOR]

Complex and large-scale distributed systems are characterized by numerous interactive data communication among distributed components over network. This paper proposes a communication data multiplexing approach as an efficient message traffic reduction scheme. This approach encodes joint outputs of sender components into a single message and applies to a distributed system involving components moving and interacting in multi-dimensional space. For performance evaluation, this paper applies uses a projectile/missile case study with realistic multi-dimensional dynamics. This paper investigates variation of system accuracy and network bandwidth requirement, while a ratio of active components and a time granule are varied. Analytical and empirical data clearly indicate the advantages of multiplexing in saving communication resources in a large-scale distributed simulation. In addition, this paper discusses effectiveness and limitation of the multiplexing approach while considering the tradeoff between system accuracy and performance. [ABSTRACT FROM AUTHOR]

Computational Grids provide an emerging highly distributed computing platform for scientific computing. Recently, service oriented architecture (SOA) is a trend of implementing software systems including Grid computing systems. SOA provides more flexibilities for Grid users at the service level. Since performance is still one of the major concerns in Grid environments, Grid service performance issues needs to be extensively investigated and studied. However, a lot of issues are still open to be explored and few work has been done on them. In this paper, we propose a Grid service monitoring architecture for flexible monitoring on various Grid services. We implemented it for monitoring WSRF (Web Service Resource Framework) services in this paper. We show how the service oriented Grid monitor work with a simple example WSRF-compliant MathService. Moreover, the relationship of the monitor and Grid super scheduler is also analyzed. In this way, the scheduler may produce service performance oriented policies that ensure optimal quality of services for Grid applications. Keywords: Grid Performance Monitoring, Grid Service Performance, Service Oriented Grid Architecture. [ABSTRACT FROM AUTHOR]

A key issue in obtaining high performance from a parallel program represented by a Directed A-cyclic Graph (DAG) is to efficiently mapping it into the target system. The problem is generally addressed in terms of task scheduling, where the tasks are the schedulable units of a program. The task scheduling problems have been shown to be NP-complete in general as well as several restricted cases. In order to be of practical use for large applications, scheduling algorithms must guarantee high performance by minimizing the schedule length and scheduling time. In this paper we propose a new task-scheduling algorithm namely, High Performance task Scheduling (HPS) algorithm for heterogeneous computing system with complexity O (v + e) (p+ log v), which provides optimal results for applications represented by DAGs. The performance of the algorithm is illustrated by comparing the schedule length, speedup, efficiency and the scheduling time with existing algorithms reported in this paper. The comparison study based on both randomly generated graphs and graphs of some real applications shows that HPS algorithm substantially outperforms existing algorithms. [ABSTRACT FROM AUTHOR]

Shared Memory is an interesting communication paradigm for SMP machines and clusters. Weak consistency models have been proposed to improve efficiency of shared memory applications. In a programming environment offering weak consistency it is a necessity to worry about individual load and store operations and about proper synchronization. In contrast to this explicit style of distributed programming hared memory systems implementing strong consistency models are easy to program and consistency is implicit. In this paper we compare two representatives: Kerrighed and Plurix implementing sequential and transactional consistency respectively. Kerrighed is a single system image operating system (OS) based on Linux whereas Plurix is a native OS for PC clusters designed for shared memory operation. The measurements presented in this paper show that strong consistency models implemented at the OS level are competitive. [ABSTRACT FROM AUTHOR]

In a preceding paper, Bruyère and the second author introduced automata, as well as rational expressions, which allow to deal with words indexed by linear orderings. A Kleene-like theorem was proved for words indexed by countable scattered linear orderings. In this paper we extend this result to languages of words indexed by all linear orderings. [ABSTRACT FROM AUTHOR]

A simple and effective way of coordinating distributed, mobile, and parallel applications is to use a virtual shared memory (VSM), such as a Linda tuple-space. In this paper, we propose a new kind of VSM, called a tagged set. Each element in the VSM is a value with an associated tag, and values are read or removed from the VSM by matching the tag. Tagged sets exhibit three properties useful for VSMs: 1Ease of use. A tagged value naturally corresponds to the notion that data has certain attributes, expressed by the tag, which can be used for later retrieval. 2Flexibility. Tags are implemented as propositional logic formulae, and selection as logical implication, so the resulting system is quite powerful. Tagged sets naturally support a variety of applications, such as shared data repositories (e.g., for media or e-mail), message passing, and publish/subscribe algorithms; they are powerful enough to encode existing VSMs, such as Linda spaces.3Security. Our notion of tags naturally corresponds to keys, or capabilities: a user may not select data in the set unless she presents a legal key or keys. Normal tags correspond to symmetric keys, and we introduce asymmetric tags that correspond to public and private key pairs. Treating tags as keys permits users to easily specify protection criteria for data at a fine granularity. This paper motivates our approach, sketches its basic theory, and places it in the context of other data management strategies. [ABSTRACT FROM AUTHOR]

Composition of a concurrent system out of components involves coordination of their mutual interactions. In component-based construction, this coordination becomes the responsibility of the glue-code language and its underlying run-time middle-ware. Reo offers an expressive glue-language for construction of coordinating component connectors out of primitive channels. In this paper we consider the problem of synthesizing Reo coordination code from a specification of a behavior as a relation on scheduled-data streams. The specification is given as a constraint automaton that describes the desired input/output behavior at the ports of the components. The main contribution in this paper is an algorithm that generates Reo code from a given constraint automaton. [ABSTRACT FROM AUTHOR]

We continue our study of the complexity of temporal logics over concurrent systems that can be described by Mazurkiewicz traces. In a previous paper (CONCUR 2003), we investigated the class of local and MSO definable temporal logics that capture all known temporal logics and we showed that the satisfiability problem for any such logic is in PSPACE (provided the dependence alphabet is fixed). In this paper, we concentrate on the uniform satisfiability problem: we consider the dependence alphabet (i.e., the architecture of the distributed system) as part of the input. We prove lower and upper bounds for the uniform satisfiability problem that depend on the number of monadic quantifier alternations present in the chosen MSO-modalities. [ABSTRACT FROM AUTHOR]

This paper studies the denotational semantics of the typed asynchronous π-calculus. We describe a simple game semantics of this language, placing it within a rich hierarchy of games models for programming languages, A key element of our account is the identification of suitable categorical structures for describing the interpretation of types and terms at an abstract level. It is based on the notion of closed Freyd category, establishing a connection between our semantics, and that of the λ-calculus. This structure is also used to define a trace operator, with which name binding is interpreted. We then show that our categorical characterization is sufficient to prove a weak soundness result. Another theme of the paper is the correspondence between justified sequences, on which our model is based, and traces in a labelled transition system in which only bound names are passed. We show that the denotations of processes are equivalent, via this correspondence, to their sets of traces. These results are used to show that the games model is fully abstract with respect to may-equivalence. [ABSTRACT FROM AUTHOR]

A new hiding scheme of digital image is introduced in this paper, which is robust to image processing such as JPEG compression, median filtering, adding Gauss noise, cropping, and histogram equalization. In this paper, we embed watermark into DWT domain, using integer modulo, retrieval of embedded watermark does not need original image. [ABSTRACT FROM AUTHOR]

In this paper, we discuss the application of a two weights neural network (TWNN) to the channel equalization problem. In particular, the purpose of the paper is to improve the previously developed TWNN equalizer with training using K-means and LMS methods; reducing the TWNN network size by considering a lesser number of TWNN kernels, and developing new techniques for determining channel order which is required to specify the structure of an TWNN equalizer. A linear regression model was used for estimating the channel order. The basic idea of reducing the network size is to select the centers, based on the channel lag. This work includes the comparison of the limits of mean square error (MSE) convergence of both a linear equalizer and an TWNN equalizer. mean square error. [ABSTRACT FROM AUTHOR]

To protect the privacy of proxy blind signature holders from dissemination of signatures by verifiers, this paper introduces universal designated-verifier proxy blind signatures. In a universal designated-verifier proxy blind signature scheme, any holder of a proxy blind signature can convert the proxy blind signature into a designated-verifier signature. Given the designated-signature, only the designated-verifier can verify that the message was signed by the proxy signer, but is unable to convince anyone else of this fact. This paper also proposes an ID-based universal designated-verifier proxy blind signature scheme from bilinear group-pairs. The proposed scheme can be used in E-commerce to protect user privacy. [ABSTRACT FROM AUTHOR]

Network forensic plays an important role in the modern network environment for computer security, but it has become a time-consuming and daunting task due to the sheer amount of data involved. This paper proposes a new method for constructing incremental fuzzy decision trees based on network service type to reduce the human intervention and time-cost, and to improve the comprehensibility of the results. At the end of paper, we discuss the performance of the forensic system and present the result of experiments. [ABSTRACT FROM AUTHOR]

Auto image segmentation can segment the image without operators interfering and is an important technique in the image processing. The Boltzmann-Color-Image-Segmentation (BCIS), which could control the degree of segmentation by adjusting the temperature parameter, is designed based on the Boltzmann-theory and the Metropolis-rule in the paper. Then the criterion function of image segmentation, which could balance between the number of segmented region and the affinity of the segmented image with the original image, is presented. Based the BCIS and Criterion function, the auto color image segmentation is schemed out by using the artificial immune algorithm. Experiments showed that the color image segmentation algorithm, which we designed in the paper, had the good capabilities. [ABSTRACT FROM AUTHOR]

This paper describes the Next Generation PARLAY X with QoS / QoE in Next Generation Network (NGN). PARLAY has introduced the architecture for the development and deployment of services by service providers over 3G network. But the existing PARLAY X does not provide the open Application Programming Interface (API) for QoS / QoE. Therefore, to solve this issue, this paper suggests the PARLAY X with QoS / QoE. The object of this paper is to support the architecture and the API of the network service for QoS / QoE in NGN. The PARLAY X can provide users with QoS / QoE in network according to the detected context such as location and speed and user's preference. The architecture of the Next Generation PARLAY X is comprised of the functions for context-awareness, adaptation, and personalization. [ABSTRACT FROM AUTHOR]

A single web service is most likely inadequate to serve the customers' business needs; it takes a selection of various web services composed together to form a business process. The cost is the primary concern of many business processes. In this paper, we propose a new solution using Genetic Algorithm (GA) in cost-reduction driven web service selection. GA is utilized to optimize business process composed of many service agents (SAg), each of which corresponds to a collection of available web services provided by multiple service providers to perform a specific function. Service selection is an optimization process with taking into account the relationships among the services. Better performance has been gotten using GA in the paper than using local service selection strategy. [ABSTRACT FROM AUTHOR]

Application of agent technology in Intrusion Detection Systems (IDSs) has been developed. Intrusion Detection (ID) agent technology can bring IDS flexibility and enhanced distributed detection capability. However, the security of the ID agent and methods of collaboration among ID agents are important problems noted by many researchers. In this paper, coordination among the intrusion detection agents by BlackBoard Architecture (BBA), which transcends into the field of distributed artificial intelligence, is introduced. A system using BBA for information sharing can easily be expanded by adding new agents and increasing the number of BlackBoard (BB) levels. Moreover the subdivided BB levels enhance the sensitivity of ID. This paper applies fuzzy logic to reduce the false positives that represent one of the core problems of IDS. ID is a complicated decision-making process, generally involving enormous factors regarding the monitored system. A fuzzy logic evaluation component, which represents a decision agent model of in distributed IDSs, considers various factors based on fuzzy logic when an intrusion behavior is analyzed. The performance obtained from the coordination of an ID agent with fuzzy logic is compared with the corresponding non-fuzzy type ID agent. [ABSTRACT FROM AUTHOR]

CBR is one of the most important artificial intelligence methods. In this paper, it is introduced to detect the variation of known attacks and to reduce the false negative rate in rule based IDS. After briefly describes the basic process of CBR and the methods of describing case and constructing case base by rules of IDS, this paper focuses on the CBR engine. A new CBR engine adapting to IDS is designed because the common CBR engines cannot deal with the specialties of intrusion cases in IDS. The structure of the new engine is described by class graph, and the core class as well as the similarity algorithm adopted by it is analyzed. At last, the results of testing the new engine on Snort are shown, and the validity of the engine is substantiated. [ABSTRACT FROM AUTHOR]

Correlation of intrusion alerts is a major technique in attack detection to build attack scenario. Rule-based and data mining methods have been used in some previous proposals to perform correlation. In this paper we integrate two complementary methods and introduce fuzzy clustering in the data mining method. To determine the fuzzy similarity coefficients, we introduce a hierarchy measurement and use weighted average to compute total similarity. This mechanism can measure the semantic distance of intrusion alerts with finer granularity than the common similarity measurement . The experimental results in this paper show that using fuzzy clustering method can reconstruct attack scenario which are wrecked by missed attacks. [ABSTRACT FROM AUTHOR]

Security and privacy are two primary concerns in RFID adoption. In this paper we focus on security issues in general and data tampering in particular. Here we present a conceptual framework to detect and identify data tampering in RFID tags. The paper surveys the existing literature and proposes to add a tamper detection component in the existing RFID middleware architecture. The tamper detection component is supported by mathematical algorithm to embed and extract secret information which can be employed to detect data tampering. [ABSTRACT FROM AUTHOR]

In the long term usage of the network, users will form certain types of habit according to their specific characteristics, individual hobbies and given restrictions. On the burst-out of worms, the overwhelming flow caused by random scanning will temporarily alter the behavior representation of users. Therefore, it is reasonable to conclude that the statistics and classification of the user habit can contribute to the detection of worms. On the basis of analysis about both users and worms, we construct the model of user-habit and propose a new approach for the early warning of worms. This paper possesses strong direction significance due to its broad applicability since extended models can be derived from the model proposed in this paper. [ABSTRACT FROM AUTHOR]

When considering information security and privacy issues most of the attention has previously focussed on data protection and the privacy of personally identifiable information (PII). What is often overlooked is consideration for the operational and transactional data. Specifically, the security and privacy protection of metadata and metastructure information of computing environments has not been factored in to most methods. Metadata, or data about data, can contain many personal details about an entity. It is subject to the same risks and malicious actions personal data is exposed to. This paper presents a new perspective for information security and privacy. It is termed Meta Privacy and is concerned with the protection and privacy of information system metadata and metastructure details. We first present a formal definition for meta privacy, and then analyse the factors that encompass and influence meta privacy. In addition, we recommend some techniques for the protection of meta privacy within the information systems. Further, the paper highlights the importance of ensuring all informational elements of information systems are adequately protected from a privacy perspective. [ABSTRACT FROM AUTHOR]

This paper presents a framework of HMM parameter adaptation technique for improving automatic speech recognition (ASR) performance in the noisy environments, which online combines the clean hidden Markov models (HMMs) with the noise model. Based on the given composite HMM corresponding to the initial recognition pass result and truncated vector Taylor series, the noise model in the cepstral domain is updated and refined using iterative Expectation-Maximization (EM) algorithm under maximum likelihood (ML) criterion. Experiments results show that the presented approach in this paper is found to greatly improve recognition performance under mismatched conditions. [ABSTRACT FROM AUTHOR]

It is important to perform the clustering task on XML documents. However, it is difficult to select the appropriate parameters' value for the clustering algorithms. Meanwhile, current clustering algorithms lack the effective mechanism to detect outliers while treating outliers as "noise". By integrating outlier detection with clustering, the paper takes a new approach for analyzing the XML documents by structure. After stating the concept of common neighbor based outlier, the paper proposes a new clustering algorithm, which stops clustering automatically by utilizing the outlier information and needs only one parameter, whose appropriate value range is decided in the outlier mining process. After discussing some features of the proposed algorithm, the paper adopts the XML dataset with different structure and other real-life datasets to compare it with other clustering algorithms. Keywords: XML Structure; Clustering; Common Neighbor. [ABSTRACT FROM AUTHOR]

In this paper, an interest region-based image retrieval system (IRBIR) that combines similarity contributions from interest regions specified by user in images to form a single value for measuring similarity between images is proposed. The interest region-based framework utilizes the segmentation result to capture the higher-level concept of images. A novel image segmentation based on Graph-Cut is proposed for the final result. The segmentation method in this paper is fast and accurate enough for the real-time image retrieval demand than previous region-based methods. Experimental and comparison results, which are performed using a general purpose database containing 2,000 images, are encouraging. [ABSTRACT FROM AUTHOR]

This paper deals with a switching PID controller using a genetic algorithm in a multi-nonlinear system. In controlling the nonlinear element of the system, there are some problems such as the limit cycle. In this study, a switching PID controller was proposed to solve problems caused by nonlinearities of system. The PID is a well-known robust controller. But, in a motor system case, it may have a limit cycle when proportional gains exceed limit. However, in other case, if the PID gain is relatively small, its torque characteristics can be too weak. In this case, the suggested switching PID controller was found to be a good approach for solving these problems despite there being difficulties in determining its boundary and gains at each boundary. In this paper, an improved genetic algorithm was used for identifying a motor system and to determining each gain for the controller. In particular, new type of crossover and mutation using a sigmoid function is applied to improve the searching ability based on the proposed improved genetic algorithm. All the processes are investigated through simulations and are verified experimentally in a real motor system. [ABSTRACT FROM AUTHOR]

This paper outlines an approach for applying acquiring numerical breast cancer image data and diagnosis using neural network algorithm in a way that is easy to classify between benign and malignance. This paper is an extended work related to our previous work [1]. In our previous work we used k-means algorithm to detect and diagnosis breast cancer tumor's region. However, to find the better results from the algorithm we need to add more numerical parameters of the breast cancer image and this algorithm has limited usage when applied with more number of parameters. Even if the cancer tumor is abnormal it was quite difficult to distinguish among those tumors. This paper summarizes the different comparative study of neural network algorithms to get the best classification of the breast cancer and explains how to acquire more numerical parameters from the breast cancer image data, so that it can help doctors to diagnosis efficiently between benign and malignance tumors. [ABSTRACT FROM AUTHOR]

Record linkage is a technique for linking records of different files that correspond to the same individual. Traditional record linkage methods needs that the files have some common variables to permit such link. In this paper we study the possibility of applying record linkage techniques when the files do not share variables. In this case we establish links based on structural information. For extracting this structural information we use in this paper twofold integrals. [ABSTRACT FROM AUTHOR]

Tracking of objects is a basic process in computer vision. This process can be formulated as exploration problems and thus can be expressed as a search into a states space based representation approach. However, these problems are hard of solving because they involve search through a high dimensional space corresponding to the possible motion and deformation of the object. In this paper, we propose a heuristic algorithm that combines three features in order to compute motion efficiently: (1) a quality of function match, (2) Kullback-Leibler measure as heuristic to guide the search process and (3) incorporation of target dynamics for computing promising search alternatives. Once target 2D motion has been calculated, the result value of the quality of function match computed is used in other heuristic algorithm with the purpose of verifying template updates. Also, a short-term memory subsystem is included with the purpose of recovering previous views of the target object. The paper includes experimental evaluations with video streams that illustrate the efficiency for real-time vision based-tasks. [ABSTRACT FROM AUTHOR]

Graphical libraries are used usually in augmentation process in order to merge together virtual and real objects. This paper focuses on a new type of virtual objects and their visual manipulation by a user. They are the captured real objects' images and are not graphically generated. We call them virtualized real objects (VRO). They are important for cases where real objects' 3D models are not available. They are also useful to test heavy or big real objects adaptation to some places. The paper presents one foundation of an augmentation that uses VRO and a camera's auto-calibration related on a 2D pattern. In the occurrence, how real objects' images can be inserted in a sequence and how to allow their manipulation for an operation of a visual disposition. [ABSTRACT FROM AUTHOR]

In this paper, we consider the possibility of obtaining a kernel machine that is sparse in feature space and smooth in output space. Smooth in output space implies that the underlying function is supposed to have continuous derivatives up to some order. Smoothness is achieved by applying a roughness penalty, a concept from the area of functional data analysis. Sparseness is taken care of by automatic relevance determination. Both are combined in a Bayesian model, which has been implemented and tested. Test results are presented in the paper. [ABSTRACT FROM AUTHOR]

This paper presents a new approach whose aim is to extent the scope of numerical models by providing them with knowledge extraction capabilities. The basic model which is considered in this paper is a multi-topographic neural network model. One of the most powerful features of this model is its generalization mechanism that allows rule extraction to be performed. The extraction of association rules is itself based on original quality measures which evaluate to what extent a numerical classification model behaves as a natural symbolic classifier such as a Galois lattice. A first experimental illustration of rule extraction on documentary data constituted by a set of patents issued form a patent database is presented. [ABSTRACT FROM AUTHOR]

The paper proposes an implementation of the agent-based population learning algorithm (PLA) within the cascade correlation (CC) learning architecture. The first step of the CC procedure uses a standard learning algorithm. It is suggested that using the agent-based PLA as such an algorithm could improve efficiency of the approach. The paper gives a short overview of both - the CC algorithm and PLA, and then explains main features of the proposed agent-based PLA implementation. The approach is evaluated experimentally. [ABSTRACT FROM AUTHOR]

The paper deals with a discrete-time recurrent neural network designed with dynamic neural models. Dynamics is reproduced within each single neuron, hence the considered network is a locally recurrent globally feed-forward. In the paper, conditions for global stability of the considered neural network are derived using the pole placement and Lyapunov second method. [ABSTRACT FROM AUTHOR]

Building an ensemble of classifiers is a useful way to improve the performance. In the case of neural networks the bibliography has centered on the use of Multilayer Feedforward (MF). However, there are other interesting networks like Radial Basis Functions (RBF) that can be used as elements of the ensemble. In a previous paper we presented results of different methods to build the ensemble of RBF. The results showed that the best method is in general the Simple Ensemble. The combination method used in that research was averaging. In this paper we present results of fourteen different combination methods for a simple ensemble of RBF. The best methods are Borda Count, Weighted Average and Majority Voting. [ABSTRACT FROM AUTHOR]

We proposed a neural network called LPPH-CSP (Lagrange Programming neural network with Polarized High-order connections for Constraint Satisfaction Problem) to solve the CSP. The CSP is a problem to find a variable assignment which satisfies all given constraints. Because the CSP has a well defined representation ability, it can represent many problems in AI compactly. From experimental results of LPPH-CSP and GENET which is a famous CSP solver, we confirmed that our method is as efficient as the GENET. In addition, unlike the other conventional CSP solvers which are discrete-valued methods, our method is a continuous-valued method and it can update all variables simultaneously, while the conventional csp solvers cannot find a solution by updating all variables simultaneously Because of the oscilation of the states. Therefore, we can expect the speed-up of LPPH-CSP if it is implemented by the hardware such as FPGA. In this paper, we extend LPPH-CSP to deal with the linear inequality constraints. By using this type of constraint, we can represent various practical problems more briefly. In this paper, we also define the CSP which has an objective function, and we extend LPPH-CSP to solve this problem. In experiment, we apply our method and OPBDP to the warehouse location problem and compare the effectiveness. [ABSTRACT FROM AUTHOR]

This paper presents a new neural network based approach to the prediction of mobile locations using signal strength measurements in a simulated metropolitan area. The prediction of a mobile location using propagation path loss (signal strength) is a very difficult and complex task. Several techniques have been proposed recently mostly based on linearized, geometrical and maximum likelihood methods. An alternative approach based on artificial neural networks is proposed in this paper which offers the advantages of increased flexibility to adapt to different environments and high speed parallel processing. The paper first gives an overview of conventional location estimation techniques and the various propagation models reported to-date, and a new signal-strength based neural network technique is then described. A simulated mobile architecture based on the COST-231 Non-line of Sight (NLOS) Walfisch-Ikegami implementation of a metropolitan environment is used to assess the generalization performance of a Multi-Layered Perceptron (MLP) Neural Network based mobile location predictor with promising initial results. [ABSTRACT FROM AUTHOR]

In this paper, a modeling method of high dimensional piecewise affine models is proposed. Because the model interpolates the outputs at the orthogonal grid points in the input space, the shape of the piecewise affine model is easily understood. The interpolation is realized by a RBFN, whose function is defined with max-min functions. By increasing the number of RBFs, the capability to express nonlinearity can be improved. In this paper, an algorithm to determine the number and locations of RBFs is proposed. [ABSTRACT FROM AUTHOR]

This paper compares the speech coder and speaker recognizer applications, showing some parallelism between them. In this paper, some approaches used for speaker recognition are applied to speech coding based on neural networks, in order to improve the prediction accuracy. Experimental results show an improvement in Segmental SNR (SEGSNR) up to 1.7 dB. [ABSTRACT FROM AUTHOR]