Search

Recently, there has been much research about cluster computing to get high performance using low-cost PCs connected with high-speed interconnection networks. In many research areas, many distributed file systems have been developed. In many distributed file systems, PVFS (Parallel Virtual File System) provides users with high bandwidth by stripping data over I/O servers. In PVFS, there is no file system cache. For a new file system cache for PVFS, we designed and implemented cooperative cache for PVFS (Coopc- PVFS). Because the previous developed Coopc-PVFS is a hint-based cooperative cache, a cache manager reads/writes files using approximately correct information so that it has a little read/write overhead. And previous studies about cooperative cache are only focused on shared read data and don't care about write performance. In this paper, we describe the design and implementation of the home-based Coopc-PVFS to improve read/write performance. Also, we evaluate and analysis the performance of the homebased Coopc-PVFS in comparison to PVFS and to the hint-based Coopc-PVFS. [ABSTRACT FROM AUTHOR]

We have previously described a hybrid keyword search that combines metadata search with a traditional keyword search over unstructured context data. This hybrid search paradigm provides the inquirer additional options to narrow the search with some semantic aspect from the XML metadata query. But in earlier work, we experienced the scalability limitations of a single-machine implementation. In this paper, we describe a scalable hybrid search on distributed databases. This scalable hybrid search provides a total query result from the collection of individual inquiries against independent data fragments distributed in a computer cluster. We demonstrate our architecture extends the scalability of a native XML query limited in a single machine and improves the performance for some queries. [ABSTRACT FROM AUTHOR]

In the different phases of a movie production, there is a need for intensive discussion about the produced sequences. Especially if the discussion members are located at different places they will have problems to precisely describe their critics for the video production. This paper describes an approach and a tool for video collaboration to simplify the video production process for video editing service providers and their customers. [ABSTRACT FROM AUTHOR]

A growing interest in Collaborative Virtual Environments (CVEs) can be observed over the last few years. Geographically dislocated users share a common virtual space as if they were at the same physical location. Although Virtual Reality (VR) is heading more and more in the direction of creating lifelike environments and stimulating all of the users senses the technology does not yet allow communication and interaction as it is in the real world. A more abstract representation is sufficient in most CVEs. This paper provides an overview on tools which can be used to enhance communication and interaction in CVEs by visualising behaviour. Not only is a set of tools presented and classified, an implementation approach on how to use these tools in a structured way in form of a framework is also given. [ABSTRACT FROM AUTHOR]

This paper introduces a tool set consisting of open source libraries that are being developed to facilitate the quick and easy implementation of collaborative VR applications. It describes functionality that can be used for generating and displaying a Virtual Environment (VE) on varied VR platforms. This is enhanced to provide collaboration support through additional modules such as networking. Two existing VR applications which make use of these tools are described. Both were developed effortlessly over a short period of time, and demonstrate the power of these tools for implementing a diverse range of applications. [ABSTRACT FROM AUTHOR]

In this paper we relate QoS management to Workflow management and discuss dependable execution on a middleware layer, named Virtual Private Grid (VPG), of sets of processes performing a workflow activity. We propose two patterns of interaction between a Workflow Management System and the distributed workflow management support. The patterns monitor resource availability & connectivity, and, in case of fault of any resource, ensure job completion by re-mapping a process graph and restarting it. We also describe current implementation of the patterns and of the run-time support. [ABSTRACT FROM AUTHOR]

The entropic lattice Boltzmann method (ELBM) has recently been shown to be capable of simulating incompressible flows with good accuracy and enhanced stability. However, the method requires that uniform grids be used, which makes it prohibitively expensive for simulating large Reynolds number flows. In this paper, an interpolation scheme is proposed to extend the applicability of this method to arbitrary non-uniform meshes, while retaining the salient features of ELBM such as stability and computational efficiency. The new scheme is used to simulate flow in a lid driven cavity as well as flow past a square cylinder and is shown to largely reduce the grid requirements. The results of the simulation agree very well with other numerical experimental data available in the literature. Keywords: Entropic Lattice Boltzmann ISLB [ABSTRACT FROM AUTHOR]

In this paper we address the problem of granting the correctness of Grid computations. We introduce a Grid Integrity Validation Scheme (GIVS) that may reveal the presence of malicious hosts by statistical sampling of computations results. Performance overheads of GIVS strategies are evaluated using statistical models and simulation. [ABSTRACT FROM AUTHOR]

A key phenomenon in the dynamics of genetic networks is the cell cycle. In the study of this phenomenon, an important task is to understand how many processes, acting on different temporal and spatial scales, interact in the cell. In this paper we deal with the problem of modelling cell cycles. We start our analysis from the Novak-Tyson model and apply this deterministic model to simulate relative protein concentrations in several different living systems, including Schixosaccharomyces pombe to validate the results. Then we generalize the model to account for the nonlinear dynamics of a cell division cycle, and in particular for special events of cell cycles. We discuss the obtained results and their implications on designing engineered regulatory genetic networks and new biological technologies. [ABSTRACT FROM AUTHOR]

In the present paper it is developed mathematical model for description of heat and mass transfer processes and predicting velocity, temperature and pollution concentrations near roadway. To describe convective transfer controlled by the wind and gravity, we use Reynolds equations for turbulent flow. The boundary value problem was solved numerically. A discrete analog for equations was obtained by means of the control volume method. Methods of predicting concentrations of automobile exhaust gases near roadways are needed for the planning and design of roads and nearby structures. [ABSTRACT FROM AUTHOR]

The purpose of this paper is to explore a time-split hybrid Galerkin scheme for the atmospheric shallow water equations. A nonlinear variant of operator integration factor splitting is employed as the time-stepping scheme. The hyperbolic system representing slow modes is discretized using the discontinuous Galerkin method. An implicit second order backward differentiation formula is applied to Coriolis and gravity wave terms. The implicit system is then discretised using a continuous Galerkin spectral element method. The advantages of such an approach include improved mass and energy conservation properties. A strong-stability preserving Runge-Kutta scheme is applied for sub-stepping. [ABSTRACT FROM AUTHOR]

This paper focuses on computational parameter identification associated with heat and mass diffusion macro-behavioral models of hydrogen storage systems from a continuum multiphysics perspective. A single wall nanotube (SWNT) based composite pellet is considered as our representative finite continuum system. The corresponding partial differential equations (PDEs) governing the spatio-temporal distribution of temperature and hydrogen concentration are formulated. Analytical solutions of the system of coupled PDEs are constructed and utilized in the context of inverse analysis. The corresponding non-linear optimization problem is formulated in order to determine the unknown parameters of the model, based on an objective function and constraints consistent with experimentally acquired data along with the physical and utilization requirements of the problem. Behavioral simulation results are presented in an effort to demonstrate the applicability of the methodology. Finally, we indicate potential extensions of this methodology to multi-scale and manufacturing process optimization. [ABSTRACT FROM AUTHOR]

In this paper, we consider subgrid modeling of a filtration flow of a fluid in a nonhomogeneous porous medium. An expression for the effective permeability coefficient for the large-scale component of the flow is derived. The permeability coefficient possesses a log-stable distribution. The obtained formulas are verified by the numerical modeling. [ABSTRACT FROM AUTHOR]

Modeling is considered an inherent part of design of Water Quality Systems. Models are used for running so-called Virtual Experiments, such as Simulation and Uncertainty Analysis. Since the complexity of VE's is constantly increasing, there is a need for techniques that allow for timely execution of complex VE's on real-world models. This paper touches one of these: distributed execution. Focus lies predominantly on the software system that was developed to this end. This system has thus far been applied to the WEST modeling and simulation software package. It is a stable, efficient and platform-independent tool for transparent distributed execution of workload on work nodes within an organization. [ABSTRACT FROM AUTHOR]

Pairwise testing is one of very practical and effective testing methods for various types of software systems. This paper proposes a novel debugging method based on pairwise testing. By analyzing the test cases and retesting with some complementary test cases, the method can narrow the factors that cause the errors in a very small range. So it can provide a very efficient and valuable guidance for the software testing and debugging. [ABSTRACT FROM AUTHOR]

In this article, we present an investigation case study based on an implementation and performance analysis of three different cache coherence strategies over a real wireless environment. Our research work considers the broadcasting timestamp, the cache coherency schema with incremental update propagation and the amnesic terminals strategies. These strategies are based on periodic broadcast of invalidation reports. The performance of these strategies is analysed through an ordinary real environment. In this environment we compare the impact of invalidation report size, broadcast interval and cache size in mobile devices. [ABSTRACT FROM AUTHOR]

We define a concept of Meta services which describes a way of abstracting and mapping a workflow to a service in computational Grid environments. By using Meta services, a workflow in a Grid environment could adapt various service concepts such as Grid services, Web services, and portal services without modification to the workflow. And the converted Meta services could be shared and reused by users. Furthermore, historical performance data could be included in the Meta service, making effective execution of the workflow possible. [ABSTRACT FROM AUTHOR]

This paper describes a Service-bAsed Grid project for Engineering computation, named SAGE. Based on the Globus toolkit, a grid-service-based architecture oriented to engineering computation is presented. To give grid users good usability, task definition and resource discovery are visually conducted. Whilst, a Quality of Service (QoS) driven user-centric scheduling strategy is proposed, two scheduling methods and steering-enabled visual interfaces are applied for different types of grid users. Result processing can be visualized in the aid of a PC-cluster and a stereopticon. The practices suggest that these mechanisms improve the convenience and QoS. [ABSTRACT FROM AUTHOR]

Campus grid storage service is to aggregate the storage resources in the servers of Campus Grid Center and colleges (or institutes, departments), and the storage resources of personal computers in the campus network. It provides storage resources registration, allocation, scheduling, and release services for users by three levels storage architecture. Due to the storage nodes' dynamites, the total storage space that nodes contribute will dynamically change with time. To study the performance of the storage service, the stochastic fluid model is adopted. By this analytical model, we got the mathematical results as follows: the function between the storage allocation probability and the number of nodes is got; if more nodes join the campus grid, the aggregated storage space will be larger, and then the available storage resources will be more; if the storage resources allocation rate is larger than the storage resources release rate, then the available storage resources will decrease. [ABSTRACT FROM AUTHOR]

There are currently many methods for triangular local interpolation: Given triangular meshes P in three dimension space, the given flat triangles are based only on the three vertices and three normal vectors, or PN triangles, construct a smooth surface that interpolates the vertices of P. In this paper a completely local interpolation scheme is presented to guarantee to join patches G1 continuously around boundary curves of each PN triangle. [ABSTRACT FROM AUTHOR]

Based on the work of Krishna and Rosenthal (1996) about combinatorial auctions bidding equilibrium analysis and Che's (1993) research about one-unit multi-attributes auctions, we construct a multi-attributes procurement combinatorial auction (MAPCA) for 2 objects, through a first-score, sealed-bid format. There are two kinds of bidders: n simple bidders and m diverse bidders considered in this model. With some assumptions, we finally obtain the equilibrium bidding strategies for the both two kinds of bidders. [ABSTRACT FROM AUTHOR]

Popov forms and weak Popov forms of matrices over noncommutative valuation domains are defined and discussed. Two new algorithms to construct these Popov forms are given, along with a description of some of their applications. [ABSTRACT FROM AUTHOR]

The investment strategies can be divided into two classes: passive investment strategies and active investment strategies. An index tracking investment strategy belongs to the class of passive investment strategies. The index tracking error and the excess return are considered as two objective functions, a bi-objective programming model is proposed for the index tracking portfolio selection problem. Furthermore, based on fuzzy decision theory, a fuzzy index tracking portfolio selection model is also proposed. A numerical example is given to illustrate the behavior of the proposed fuzzy index tracking portfolio selection model. [ABSTRACT FROM AUTHOR]

Spatial index is very important in GIS. PK tree and Hilbert R tree are two well-known spatial index structures. Comparison operations are very little in PK tree, while disk I/O operations are quite little in Hilbert R tree. PK+ Tree is an improved spatial index structure from PK tree. In PK+ tree, Comparison operations are less than in Hilbert R tree, while disk I/O operations are almost the same as in Hilbert R tree. [ABSTRACT FROM AUTHOR]

We numerically simulate effects of the self-heating on the current-voltage characteristics of InGaP/GaAs heterojunction bipolar transistors (HBTs). A set of coupled nonlinear ordinary differential equations (ODEs) of the equivalent circuit of HBT is formed and solved numerically in the large-signal time domain. We decouple the corresponding ODEs using the waveform relaxation method and solve them with the monotone iterative method. The temperature-dependent energy band gap, the current gain, the saturation current, and the thermal conductivity are considered in the model formulation. The power-added efficiency and the 1-dB compression point of a three-finger HBT are calculated. This approach successfully explores the self-heating and the thermal coupling phenomena of the three-finger transistors under high power and high frequency conditions. The numerical algorithm reported here can be incorporated into electronic computer-aided design software to simulate ultra-large scale integrated and radio frequency circuits. [ABSTRACT FROM AUTHOR]

A correctly and continually working authorization and authentication service is essential for the grid computing system, so it is very necessary to maintain efficient this service with high availability and integrity in the face of a variety of attacks. An intrusion-resilient framework of authorization and authentication service for grid computing system is presented in this paper. This service is able to provide fault tolerance and security even in the presence of a fraction of corrupted authorization and authentication servers, avoiding any single point of failure. We use a cryptographic (f, n) secret sharing scheme to distribute parts of the clients' proxy certificates and use a secure multi-party computation scheme to perform the signatures such that the proxy certificate can be issued in a distributed fashion without reassembly when a legal client registrant at the Globus host. By using Non-Malleable Proof, the "man-in-the-middle attack" can be prevented; by distributing the secret data across several authorization and authentication servers, the compromise of a few servers will not compromise the availability of data. And, under the assumption of a Diffie-Hellman decisional problem, a passive adversary gets zero knowledge about the system's private key X, and so cannot to issue the certification for any client, neither to impersonate a legal authorization and authentication server. [ABSTRACT FROM AUTHOR]

Advance Reservations are an important concept to support QoS and Workflow Scheduling in Grid environments. However, the impact of reservations from the Grid on the performance of local schedulers is not yet known. Using discrete event simulations we evaluate the impact of reservations on planning-based resource management of standard batch jobs. Our simulations are based on a real trace from the parallel workload archive. By introducing a new option for scheduling reservations in planning-based resource management, less reservation requests are rejected. Our results are important for increasing the acceptability of the Grid technology. We show, that a limited number of additional resource reservations from the Grid have only a limited impact on the performance of the traditionally submitted batch jobs. [ABSTRACT FROM AUTHOR]

The article presents application of artificial immune algorithms in classification of vectorized medical data sets. Artificial immune network was created and trained for the purpose of arterial blood gasometry parameters (pH, pCO2, pO2, HCO3) classification. Training data originates from the Infant Intensive Care Unit of the Polish - American Institute of Pediatry, Collegium Medicum, Jagiellonian University in Cracow. [ABSTRACT FROM AUTHOR]

In just the past few years, the field of mathematical modeling with equation-based tools has seen a number of advances and innovations that allow software developers to create programs that are at once far more powerful yet easier to use. To illustrate the benefits that users gain from the latest features and capabilities, this paper examines a problem where software that sup ports extended multiphysics calculates the applied voltage needed to produce a predefined thermally induced bending of a microscale device. The package used in this problem is FEMLAB, an equation-based environment for solving a large class of systems of coupled partial differential equations. [ABSTRACT FROM AUTHOR]

Tourist beaches on the southern coast of Turkey are surveyed in order to facilitate a standardised fuzzy approach to be used in litter prediction and to assess the aesthetic state of the coastal environment for monitoring programs. During these surveys the number of litter items on beaches were counted and recorded in different categories. The main source of litter on beaches was determined as "beach users". A fuzzy system was developed to predict the classification of the beaches, since uncertainty was generally inherent in beach work due to the high variability of beach characteristics and the sources of litter categories. This resulted in effective utilization of "the judgment and knowledge of beach users" in the evaluation of beach gradings. [ABSTRACT FROM AUTHOR]

Replication of content on geographically distributed servers can improve both performance and reliability of the Web service. Anycast is a one-to-one-of-many delivery technique that allows a client to choose a content server of a set of replicated servers. Presenting numerous mirror content servers to a client implies the difficult problem of finding the best server in terms of network performance. We formulate an optimization problem of anycast flows assignment in a connection-oriented network. This is a 0/1, NP-complete problem, which is computationally very difficult due to the size of solution space and constraints. Therefore, we propose computationally effective heuristic algorithm. To our knowledge, this is the first proposal to solve the problem of anycast flow assignment in connection-oriented network. Results of simulations are shown to evaluate performance of proposed algorithm for various scenarios. [ABSTRACT FROM AUTHOR]

We discuss the use of the computer algebra system Maple ® to analyze the heat transport when a heated sphere melts its way through a solid medium. The combined power of symbolic computation and numerical simulation available in Maple is used to analyze a model in which the molten layer around the moving sphere is augmented by a surrounding thermal-layer. In particular, we carefully study the effects of imposing a variety of boundary conditions. [ABSTRACT FROM AUTHOR]

Cluster configurations are a cost effective scenarios which are becoming common options to enhance several classes of applications in many organizations. In this article, we present a research work to enhance the load balancing, on dedicated and non-dedicated cluster configurations, based on a genetic machine learning algorithm. Classifier systems are learning machine algorithms, based on high adaptable genetic algorithms. We developed a software package which was designed to test the proposed scheme in a master-slave Cow and Now environment. Experimental results, from two different operating systems, indicate the enhanced capability of our load balancing approach to adapt in cluster configurations. [ABSTRACT FROM AUTHOR]

We propose a space-efficient key pre-distribution scheme based on quasi-orthogonal finite projective plane. This approach, compared to the previous approaches, guarantees full connectivity and the uniqueness of pairwise keys. Moreover, the size of the key ring depends not on the size of the key pool, but on the size of the network. The actual order of the key ring size is only $O(\sqrt{N})$, where N is the size of the network. As a result, our approach provides better scalability than previous approaches. [ABSTRACT FROM AUTHOR]

Sequence Pair (SP) is an elegant representation for the block placement of IC Design, and it is usually imperative to generate the SP from an existing placement. A quick generation method and one concise algorithm are proposed instead of the original unfeasible one. It is also shown that if the relations of any two blocks are either vertical or horizontal, the solution space size of a representation is (n!)2 if it is P*-admissible. The analytical and experimental results of the algorithm both show its superiority in running time. [ABSTRACT FROM AUTHOR]

We present a new neural network model for classification of facial expressions based on dimension model that is illumination-invariant and without detectable cues such as a neutral expression. The neural network model on the two-dimensional structure of emotion have improved the limitation of expression recognition based on a small number of discrete categories of emotional expressions, lighting sensitivity, and dependence on cues such as a neutral expression. [ABSTRACT FROM AUTHOR]

In order to remove neighboring abdominal organs of the liver, we propose an improved fully automatic liver segmentation using histogram tail threshold (HTT) algorithms. A region of interest of the liver is first segmented. A left HTT (LHTT) algorithm is performed to eliminate the pancreas, spleen, and left kidney. After the right kidney is eliminated by the right HTT (RHTT) algorithm, the robust liver structure is segmented. From the results of experiments, the improved automatic liver segmentation using HTT algorithms has strong similarity performance as manual segmentation by medical doctor. [ABSTRACT FROM AUTHOR]

We propose a novel technique of multimodality volume fusion using graphics hardware that solves the depth cueing problem with less time consumption. Our method consists of three steps. First, it takes two volumes and generates sample planes orthogonal to the viewing direction following 3D texture mapping volume rendering. Second, it composites textured slices each from different modalities with several compositing operations. Third, alpha blending for all the slices is performed. For the efficient volume fusion, a pixel program is written in HLSL(High Level Shader Language). Experimental results show that our hardware-accelerated method distinguishes the depth of overlapping region of the volume and renders them much faster than conventional ones on software. [ABSTRACT FROM AUTHOR]

The Soboĺ sequence is the most popular quasirandom sequence because of its simplicity and efficiency in implementation. We summarize aspects of the scrambling technique applied to Soboĺ sequences and propose a new simpler modified scrambling algorithm, called the multi-digit scrambling scheme. Most proposed scrambling methods randomize a single digit at each iteration. In contrast, our multi-digit scrambling scheme randomizes one point at each iteration, and therefore is more efficient. After the scrambled Soboĺ sequence is produced, we use this sequence to evaluate a particular derivative security, and found that when this sequence is numerically tested, it is shown empirically to be far superior to the original unscrambled sequence. [ABSTRACT FROM AUTHOR]

We describe the Agents Channelling ContExt Sensitive Services (ACCESS) architecture, an open agent-based architecture that supports the development and deployment of multiple heterogeneous context sensitive services. We detail the ACCESS architecture and describe the scenario of an individual arriving in a city and using his ACCESS enabled PDA to secure lodgings. [ABSTRACT FROM AUTHOR]

To enhance scalability of replicated services a large number of mobile agents attempt to access in mobile agent systems, we present a new strategy to apply an appropriate passive replication protocol to each replicated service according to its execution behavior because deterministic services require weaker constraints to ensure their consistency than non-deterministic ones. For this goal, two passive replication protocols are introduced for non-deterministic services and for deterministic services respectively. They both allow visiting mobile agents to be forwarded to and execute on any node performing a service agent, not necessarily the primary agent. Especially, in case of the protocol for deterministic services, after a backup service agent has received a mobile agent request and obtained the delivery sequence number of the request from the primary service agent, the backup agent, not the primary one, is responsible for processing the request and coordinating with the other replica service agents. [ABSTRACT FROM AUTHOR]

In biological systems the process of symbol grounding is realised at a neural level. But in case of artificial subjects it must be done by certain organisations of private data structures and data collections, and by internal data processing and recognition algorithms. In this work a possible way of implementation of the well - known problem of symbol grounding is presented. The idea of an algorithm of grounding formulas in a private agent's data structure is given. It is assumed that from the agent's point of view stored perceptions are the original source of any meaning accessible for cognitive processes of symbol grounding. This algorithm is used when the state of a particular object from external world is not known. As a results the logic formulas with modal operators and descriptive connectives are generated. [ABSTRACT FROM AUTHOR]

We present the extended notion of pure exchange economy under uncertainty, called an economy with awareness structure, where each trader having a strictly monotone preference makes decision under his/her awareness and belief, and we introduce a generalized notion of equilibrium for the economy, called an expectations equilibrium in awareness. We show the existence theorem of the equilibrium and the fundamental welfare theorem for the economy, i.e., an allocation in the economy is ex-ante Pareto optimal if and only if it is an expectations equilibrium allocation in awareness. [ABSTRACT FROM AUTHOR]

An interesting topic in dynamical systems is chaotic synchronization, that is, the possibility to synchronize the behavoir of several chaotic systems. Among the variety of available schemes, those dealing with some sort of control mechanisms have received increasing attention in recent years. In this work, we applied the program Mathematica to study the control strategies of chaotic synchronization. In our opinion, the powerful symbolic, numeric and graphic capabilities of Mathematica make this software an ideal tool to analyze the problem on hand. [ABSTRACT FROM AUTHOR]

We present newly implemented procedures in SyNRAC, which is a Maple package for solving real algebraic constraints derived from various engineering problems. The current version of SyNRAC has added quantifier elimination (QE) by cylindrical algebraic decomposition (CAD) as well as QE by virtual substitution for weakly parametric linear formulas. We also show an application of CAD-based QE to the common Lyapunov function problem. [ABSTRACT FROM AUTHOR]

The physics laws, which govern the atmospheric phenomena, are mostly non-linear and therefore the application of the conventional Fourier spectral analysis on the time series of the atmospheric quantities reveals that these are usually non-stationary. Quite often these non-stationarities conceals the existing correlations and therefore new analytical techniques capable to eliminate non-stationarities in the data should be employed. The most recent analytical methods used along these lines are the wavelet techniques and the detrended fluctuation analysis. Much attention has been paid recently to the latter technique, which has already proved its usefulness in a large variety of complex systems. As a paradigm, the detrended fluctuation analysis is applied to the column ozone data. Specifically the zonally and globally averaged column ozone observations conducted by ground-based (1964-2004) and satellite-borne (1979-2003) instrumentation are employed to detect long-range correlations in column ozone time series. The results show that column ozone fluctuations exhibit persistent long-range power-law correlations for all time lags between 4 months - 11 years. [ABSTRACT FROM AUTHOR]

We have presented a principal component analysis linear-weighted model (PCALWM) for credit scoring in [5,6], this article is a further study on this model. We revised the application procedure in the credit scoring, and tested it by a larger real-life credit card dataset. In comparison with some well-known scores, the empirical results of the PCALWM can achieve a favorable KS distance. The study on some application features of this model in the credit decision-making shows that the model can help the credit issuers to select the best trade-off among the enterprise stratagem, marketing and credit risk management. [ABSTRACT FROM AUTHOR]

This study proposes a novel forecasting approach - an adaptive smoothing neural network (ASNN) - to predict foreign exchange rates. In this new model, adaptive smoothing techniques are used to adjust the neural network learning parameters automatically by tracking signals under dynamic varying environments. The ASNN model can make the network training process and convergence speed faster, and make network's generalization stronger than the traditional multi-layer feed-forward network (MLFN) model does. To verify the effectiveness of the proposed model, three major international currencies (British pounds, euros and Japanese yen) are chosen as the forecasting targets. Empirical analyses reveal that the proposed novel forecasting model outperforms the other comparable models. Furthermore, experimental results also show that the proposed model is an effective alternative approach for foreign exchange rate forecasting. [ABSTRACT FROM AUTHOR]

A practical component-based WebGIS named as Geo-Union is presented. Geo-Union consists of four layers: storage layer, service layer, component layer and application layer. Service layer can be partitioned into another two layers: client service layer and server service layer. The architectures and object constitutions of each layer in Geo-Union are discussed in details. The Web application model of Geo-Union is also presented. At last, some future works in WebGIS, such as interoperability, security, distributed computing and intelligent computing, are indicated and simply explored. [ABSTRACT FROM AUTHOR]

In this article we report on our efforts to test and expand the current state-of-the-art in eigenvalue solvers applied to the field of nanotechnology. We singled out the nonlinear conjugate gradients (CG) methods as the backbone of our efforts for their previous success in predicting the electronic properties of large nanostructures and made a library of three different solvers (two recent and one new) that we integrated into the parallel PESCAN (Parallel Energy SCAN) code [3] to perform a comparison. [ABSTRACT FROM AUTHOR]