|A Cognitive Framework for Knowledge-based Process Design|
Hans Van Leijen, Walter R.J. Baets
We propose a framework for redesigning knowledge-intensive business processes that is inspired by the knowledge-based theory of the firm, and based on ideas from cognitive science. It views a business process as a problem-solving task consisting of five phases: recognition, decomposition, planning, action and evaluation. The coordination of these tasks among multiple agents is viewed as distributed cognition. We give some general principles for identifying process improvements based on manipulating these phases.
A Neural Network Model for Urban Traffic Volumes Compression
Xiaoling Ou, Yi Zhang, Jiangtao Ren, Danya Yao
Traffic data are the information source for traffic control and management. With the development and integration of Intelligent Transportation Systems, many applications and their respective sensors and detectors are a rich source of data about transportation system characteristics and performance. However, because of the limitation of databases and devices, the huge amounts of traffic data can not be stored without reduction. In this paper, an approach for urban traffic volume compression based on artificial neural network is proposed. The lossy compression of data is realized by using a set of three-layer back-propagation neural networks to remove the correlation within traffic volumes. The model has both a small reproduction error and a relatively high compression ratio.
A New Multi-Layered Fuzzy Image Filter for Removing Impulse Noise
Russel J Stonier, Mohamed M Anver
In this paper we develop a fuzzy image .lter which consists of a multi-layered fuzzy structure based on the weighted fuzzy blend filter for the removal of noise from images heavily corrupted by impulse noise, while preserving the intricate details of the image. The introduction of multi-layered fuzzy systems substantially decreases the number of rules to be learnt. We then show how Evolutionary Algorithms (EAs) can be used to effectively learn the fuzzy rules in each knowledge base. Results are presented for impulse noise corruption of the well-known ‘Lena’ image.
Modeling Time Series Signal Patterns by Statistical Distribution of Prediction Errors and its Application to Speaker Identification
Qian-Rong Gu, Tadashi Shibata
A new method that uses statistical distribution of prediction error vectors to build models of time series patterns has been developed. A universal predictor is firstly established from universal training data. Then, properties common to all the patterns are removed from the training data by the predictor. The residuals, i.e., the prediction errors, hold the characteristics of individual patterns. After clustering the prediction errors to a universal codebook, the predictor and the codebook are applied to individual training data sets to obtain the usage histograms of code vectors in the universal codebook, namely, the statistical distribution of prediction error vectors. These histograms represent the properties of individual patterns, and can be used as models in pattern recognition applications. This method is not restricted to any specific signals. As a demonstration, we utilized it to speaker identification application. It performed as well as other modeling methods under the text-dependent condition.
Multistable Phase Locking Patterns in a Bursting Neural System
Seon-Hee Park, Young-Sup Hwang, Jae-Hun Choi
We quantitatively analyze the multistability of dynamic patterns of a busting neural system with diffusive coupling. Through effective coupling analysis, we show that the system is not in-phase locking but exhibits various phase locking patterns, each of which corresponds to the stable fixed points of the effective coupling. The simulation proves the validity of the effective coupling method in analyzing the multistability of neural systems which presents complicated dynamic patterns such as bursting.
New Photonic System for Optical Packet Switching
F. Rudge Barbosa, A. Campos Sachs, M. Tosi Furtado
Fast optical switching (ms timebase) is realized by using a RF frequency tone inserted in the optical packet that carries a digital payload. By using a highly selective RF filtering for optical packet header frequency recognition, we have obtained excellent performance in optical switching function.. The RF header is detected at optical node input, and signals the node switching control, which instantly directs the packet to a prescribed output. No electronic processing of the digital payload is performed. The optical circuit is noise-free, has very low crosstalk, and is extremely selective in header frequency detection. BER measurements for payload consistently yield figures as low as 10-12 . This system is applicable to optical metropolitan and access networks, and is fully compatible with DWDM systems.
On Neural-Net Based Variable Structure Multiple Model Method: Neural-Net Design
Daebum Choi, Monica Samal, Byungha Ahn
In order to track a maneuvering target, multiple model (MM) methods have been researched. Almost MM algorithms have been developed based on Markov process. However, Markov based MM method is difficult to design and application-dependent. To solve this problem, Daebum Choi, et al proposed basic idea of neural-net based VSMM . In this paper, we will show the design procedure of neural-net based MM and discuss how does it work in details.
Optimal boundary control of a tracking problem for a parabolic distributed system with open-loop control using evolutionary algorithms
Russel J Stonier, Michael J Drumm
In this paper we examine the application of evolutionary algorithms to find open-loop control solutions of the optimal control problem arising from the semidiscretisation of a linear parabolic tracking problem with boundary control. The solution is compared with the solutions obtained by methods based upon the variational equations of the Minimum Principle and the finite element method.
Optimal Wavelets for Speech Signal Representations
Shonda L. Walker, Simon Y. Foo
It is well known that in many speech processing applications, speech signals are characterized by their voiced and unvoiced components. Voiced speech components contain dense frequency spectrum with many harmonics. The periodic or semi-periodic nature of voiced signals lends itself to Fourier Processing. Unvoiced speech contains many high frequency components and thus resembles random noise. Several methods for voiced and unvoiced speech representations that utilize wavelet processing have been developed. These methods seek to improve the accuracy of wavelet-based speech signal representations using adaptive wavelet techniques, superwavelets, which uses a linear combination of adaptive wavelets, gaussian methods and a multi-resolution sinusoidal transform approach to mention a few. This paper addresses the relative performance of these wavelet methods and evaluates the usefulness of wavelet processing in speech signal representations. In addition, this paper will also address some of the hardware considerations for the wavelet methods presented.
|Random Shuffling Permutations of Nucleotides|
Shiquan Wu, Xun Gu
In this paper, we discuss a shuffling sequence problem: Given a DNA sequence, we generate a random sequence that preserves the frequencies of all mononucleotides, dinucleotides, trinucleotides or some high order base-compositions of the given sequence. Two quadratic running time algorithms, called Frequency-Counting algorithm and Decomposition- and-Reassemble algorithm, are presented for solving the problem. The first one is to count all frequencies of the mononucleotides, dinucleotides, trininucleotides, and any high order base-compositions in the given sequence. The second one is to generate a random DNA sequence that preserves the mononucleotides, dinucleotides, trinucleotides, or some high order base- compositions. The two algorithms are implemented into a program ShuffleSeq (in C) and is available at http://www.cs.iastate.edu/~ sqwu/ShuffleSeq.html.
World Wide Web Usage Mining Systems and Technologies
Wen-Chen Hu, Xuli Zong, Chung-wei Lee, Jyh-haw Yeh
Web usage mining is used to discover interesting user navigation patterns and can be applied to many real-world problems, such as improving Web sites/pages, making additional topic or product recommendations, user/customer behavior studies, etc. This article provides a survey and analysis of current Web usage mining systems and technologies. A Web usage mining system performs five major tasks: i) data gathering, ii) data preparation, iii) navigation pattern discovery, iv) pattern analysis and visualization, and v) pattern applications. Each task is explained in detail and its related technologies are introduced. A list of major research systems and projects concerning Web usage mining is also presented, and a summary of Web usage mining is given in the last section.
A Software Architecture for Control of Value Production in Federated Systems
Jay S. Bayne
Federated enterprises are defined as interactive commercial entities that produce products and consume resources through a network of open, free-market transactions. Value production in such entities is defined as the real-time computation of enterprise value propositions. These computations are increasingly taking place in a grid-connected space – a space that must provide for secure, real-time, reliable end-to-end transactions governed by formal trading protocols. We present the concept of a value production unit (VPU) as a key element of federated trading systems, and a software architecture for automation and control of federations of such VPUs.
An Alternative Sorting Procedure for Interactive Group Decision Support Based on the Pseudo-criterion Concept
Andrej Bregar, József Gyorkos, Ivan Rozman
An original interactive procedure is proposed, which aims at overcoming some of the major weaknesses of existing pseudocriterion based methods for group decision analysis. It refers to absolute judgements of feasible alternatives and is focused on complementary activities of opinion elicitation and robustness analysis. As a foundation, four interdependent principles are introduced – problem localization, interactivity on the basis of progressiveness approach, semiautomatic derivation of criteria weights according to selective effects of veto thresholds, and group consensus seeking. The principles are grounded and realized by appropriate methodological solutions.
A Recurrent Neural Network Approach to Rear Vehicle Detection Which Considered State Dependency
Kayichirou Inagaki, Shozo Sato, Taizo Umezaki
Experimental vision-based detection often fails in cases when the acquired image quality is reduced by changing optical environments. In addition, the shape of vehicles in images that are taken from vision sensors change due to approaches by vehicle. Vehicle detection methods are required to perform successfully under these conditions. However, the conventional methods do not consider especially in rapidly varying by brightness conditions. We suggest a new detection method that compensates for those conditions in monocular vision-based vehicle detection. The suggested method employs a Recurrent Neural Network (RNN), which has been applied for spatiotemporal processing. The RNN is able to respond to consecutive scenes involving the target vehicle and can track the movements of the target by the effect of the past network states. The suggested method has a particularly beneficial effect in environments with sudden, extreme variations such as bright sunlight and shield. Finally, we demonstrate effectiveness by state-dependent of the RNN-based method by comparing its detection results with those of a Multi Layered Perceptron (MLP).
Artery: A Data-Centric Architecture for Wireless Sensor Networks
Lan Lin, Hailin Wu
Sensor networks, composed of large amount of micro-sensors, are considered promising, both in academic research and in real life applications. To ensure highly efficient communications between event observers and sensor network users, new infrastructures and algorithms are being developed. This paper describes Artery, a novel architecture that delivers queries and data between multiple observers and multiple mobile users. Simulation results show that Artery outperforms some major data dissemination algorithms.
Virtual AM Stereo and Surround Sound to setup AM/FM Radio Theatre
Selvakumaran Vadivelmurugan, Veeraraghavan Krishnan
Introduction of virtual surround sound and stereo to AM radio has been proposed in this study. This technology can be further applied to aid the construction of an AM radio theatre. Adding to the advantages of AM, the lower bandwidth, higher range and simpler circuitry, AM can now offer excellent sound effect with the post-transmission process. The motivation for the introduction of virtual surround sound is the poor quality of AM sound. In this study, the response by human ear has been thoroughly investigated and the methodology to create virtual surround sound has been developed. The elements essential to setup audio theatre such as the components of audio chain, multiple unit audio speaker, inner section of the ear, psychological effect of different ranges of frequency and radio theatre design have been extensively studied on the basis of Helmholtz audition theory. The vital changes include the different frequency division multiplexing of message at the transmitting end, three phases of the process, resulting in the vertical and horizontal digital connection, espresso program and the 3x12 speaker design theatre.
Secure Wireless Sensor Networks: Problems and Solutions
Fei Hu, Jim Ziobro, Jason Tillett, Neeraj K. Sharma
As sensor networks edge closer towards wide-spread deployment, security issues become a central concern. So far, the main research focus has been on making sensor networks feasible and useful, and less emphasis was placed on security. This paper analyzes security challenges in wireless sensor networks and summarizes key issues that should be solved for achieving the ad hoc security. It gives an overview of the current state of solutions on such key issues as secure routing, prevention of denial-of-service and key management service. We also present some secure methods to achieve security in wireless sensor networks. Finally we present our integrated approach to securing sensor networks.
Two Dimensional Projection Pursuit Applied to Gaussian Mixture Model Fitting
Natella Likhterov, Mayer Aladjem
In this paper we seek a Gaussian mixture model (GMM) of an n-variate probability density function. Usually the parameters of GMMs are determined by a maximum likelihood (ML) criterion. A practical deficiency of ML fitting of GMMs is poor performance when dealing with high-dimensional data since a large sample size is needed to match the accuracy that is possible in low dimensions. We propose a method to fit the GMM to multivariate data which is based on the two-dimensional projection pursuit (PP) method. By means of simulations we compare the proposed method with a one-dimensional PP method for GMM. We conclude that a combination of one- and twodimensional PP methods could be useful in some applications.