Moment analysis for load-dependent mixed product form queueingnetworks

A solution algorithm is developed for higher moments of the number of jobs at the stations in load-dependent mixed product form queuing networks. Such networks are important because they can be used to model computer systems and computer communication networks. A relationship is proved and applied between the derivatives of these moments and higher moments. The derivative measures are calculated using a mean value analysis type algorithm called moment analysis. The authors describe the model and review the product form solutions for load-dependent mixed queuing networks. They develop the moment analysis algorithm and the formulas for the performance measures such as the response time, throughput and mean number of jobs in the stations. The complexity of the algorithm is discussed     

The use of noise properties in set theoretic estimation

In most digital signal processing problems, the goal is to estimate an object from noise corrupted observations of a physical system. The authors describe how a wide range of probabilistic information pertaining to the noise process can be used in a general set theoretic estimation framework. The basic principle is to constrain the sample statistics of the estimation residual to be consistent with those probabilistic properties of the noise which are available and to construct sets accordingly in the solution space. Adding these sets to the collection of sets describing the solution will yield a smaller feasibility set and, hence, more reliable estimates. Pieces of information relative to quantities such as range, moments, absolute moments, and second and higher order probabilistic attributes are considered, and properties of the corresponding sets are established. Simulations are provided to illustrate the theoretical developments     

基于原点矩偏导的K分布杂波参数估计

K分布的参数估计研究对于雷达杂波特性预测和估计具有重要意义。基于矩估计的K分布参数估计方法通过联立不同阶原点矩进行参数求解,这些不同阶数下的联合,在数据长度受限情况下会产生误差。因此,通过计算原点矩偏导和原点矩之间的关系,推导出了一种新的K分布杂波参数估计方法。该方法在同阶原点矩的条件下进行参数估计,避免了不同阶原点矩之间的估计误差,具有更好的估计性能。通过仿真和实测杂波数据,分析比较了该方法与其他矩估计法的参数估计有效率和估计精度,该方法具有100%的估计有效率且估计精度更高。高阶矩对数据敏感,在矩估计法中应尽量选取低阶矩,通过合理选取阶数k可以得到较为理想的估计结果和精度。     

Wavelet image extension for analysis and classification ofinfarcted myocardial tissue

Some computer applications for tissue characterization in medicine and biology, such as analysis of the myocardium or cancer recognition, operate with tissue samples taken from very small areas of interest. In order to perform texture characterization in such an application, only a few texture operators can be employed: the operators should be insensitive to noise and image distortion and yet be reliable in order to estimate texture quality from the small number of image points available. In order to describe the quality of infarcted myocardial tissue, the authors propose a new wavelet-based approach for analysis and classification of texture samples with small dimensions. The main idea of this method is to decompose the given image with a filter bank derived from an orthonormal wavelet basis and to form an image approximation with higher resolution. Texture energy measures calculated at each output of the filter bank as well as energies of synthesized images are used as texture features in a classification procedure. The authors propose an unsupervised classification technique based on a modified statistical t-test. The method is tested with clinical data, and the classification results obtained are very promising. The performance of the new method is compared with the performance of several other transform-based methods. The new algorithm has advantages in classification of small and noisy input samples, and it represents a step toward structural analysis of weak textures     

位积分方程组矩量法用于精确计算非均匀介质柱的电磁散射

位积分方程组的主要特点是以电磁位为未知函数,这些未知函数在具有不同电磁参数的介质分界面处是连续的,因而在矩量法的实现过程中能够非常方便地应用高阶插值基函数来展开未知函数,以便获得高精度的解。但是,经典的点匹配方案使该模型的数值稳定性较差。本文用位积分方程组矩量法模型计算任意截面非均匀介质柱的电磁散射,采用三角形离散方案和高阶插值基函数,在测试过程中应用新提出的测试方法,克服了原位方程组矩量法模型的数值不稳定性。对矩量法矩阵中自阻抗元素的奇异性处理方法也作了详细介绍。文中提供的数值结果表明,该方法是精确、稳定的。     

Low dimensional adaptive texture feature vectors from class distance and class difference matrices

In many popular texture analysis methods, second or higher order statistics on the relation between pixel gray level values are stored in matrices. A high dimensional vector of predefined, nonadaptive features is then extracted from these matrices. Identifying a few consistently valuable features is important, as it improves classification reliability and enhances our understanding of the phenomena that we are modeling. Whatever sophisticated selection algorithm we use, there is a risk of selecting purely coincidental "good" feature sets, especially if we have a large number of features to choose from and the available data set is limited. In a unified approach to statistical texture feature extraction, we have used class distance and class difference matrices to obtain low dimensional adaptive feature vectors for texture classification. We have applied this approach to four relevant texture analysis methods. The new adaptive features outperformed the classical features when applied to the most difficult set of 45 Brodatz texture pairs. Class distance and difference matrices also clearly illustrated the difference in texture between cell nucleus images from two different prognostic classes of early ovarian cancer. For each of the texture analysis methods, one adaptive feature contained most of the discriminatory power of the method.     

Mixed complete Padé model reduction: a useful formulation for closed loop design

A simple formulation for derivation of reduced order models for linear time invariant systems by mixed complete Padé approximation is presented. Various reduced order models of the same order with different combinations of the number of matched Markov parameters and time moments can be derived by equations that resemble the simple Padé equations for approximation at a single point. The advantages of obtaining models by a complete Padé approximation is emphasised.     

Electromagnetic radiation and scattering from finite conducting anddielectric structures: surface/surface formulation

An efficient and accurate numerical procedure for the analysis of the electromagnetic scattering and radiation from arbitrarily shaped, composite finite conducting and dielectric bodies is proposed. A set of coupled electric field integral equations involving surface equivalent electric and magnetic currents is used. The coupled integral equations are solved through planar triangular patch modeling and the method of moments. Two separate, mutually orthogonal vector functions for each edge connecting a pair of triangular patches have been developed. Numerical results for disk/cone and cylinder/cone structures are compared with other available data. Limited comparison with experimental data has also been made     

Efficient maximum likelihood classification for imagingspectrometer data sets

A simplified maximum likelihood classification technique for handling remotely sensed image data is proposed which reduces, significantly, the processing time associated with traditional maximum likelihood classification when applied to imaging spectrometer data, and copes with the training of geographically small classes. Several wavelength subgroups are formed from the complete set of spectral bands in the data, based on properties of the global correlation among the bands. Discriminant values are computed for each subgroup separately and the sum of discriminants is used for pixel labeling. Several subgrouping methods are investigated and the results show that a compromise among classification accuracy, processing time, and available training pixels can be achieved by using appropriate subgroup sizes     

Neural network approach to land cover mapping

A pattern classification method is proposed for remote sensing data using neural networks. First, the authors apply the error backpropagation (BP) algorithm to classify the remote sensing data. In this case, the classification performance depends on a training data set. In order to get stable and precise classification results, the training data set is selected based on geographical information and Kohonen's self-organizing feature map. Using the training data set and the error backpropagation algorithm, a layered neural network is trained such that the training patterns are classified with a specified accuracy. After training the neural network, some pixels are deleted from the original training data set if they are incorrectly classified and a new training data set is built up. Once training is complete, a testing data set is classified by using the trained neural network. The classification results of LANDSAT TM data show that this approach produces excellent results which are more realistic and noiseless compared with a conventional Bayesian method     

Image set-based classification using collaborative exemplars representation

In many classification tasks, multiple images that form image set may be available rather than a single image for object. For image set classification, crucial issues include how to simply and efficiently represent the image sets and deal with outliers. In this paper, we develop a novel method, called image set-based classification using collaborative exemplars representation, which can achieve the data compression by finding exemplars that have a clear physical meaning and remove the outliers that will significantly degrade the classification performance. Specifically, for each gallery set, we explicitly select its exemplars that can appropriately describe this image set. For probe set, we can represent it collaboratively over all the gallery sets formed by exemplars. The distance between the query set and each gallery set can then be evaluated for classification after resolving representation coefficients. Experimental results show that our method outperforms the state-of-the-art methods on three public face datasets, while for object classification, our result is very close to the best result.     

Electromagnetic scattering from arbitrary shaped conducting bodiescoated with lossy materials of arbitrary thickness

A simple and efficient numerical technique is presented to solve the electromagnetic scattering problem of coated conducting bodies of arbitrary shape. The surface equivalence principle is used to formulate the problem in terms of a set of coupled integral equations involving equivalent electric and magnetic surface currents which represent boundary fields. The conducting structures and the dielectric materials are modeled by planar triangular patches, and the method of moments is used to solve the integral equations. Numerical results for scattering cross sections are given for various structures and compared with other available data. These results are proved accurate by a number of representative examples     

Optimizing feature extraction for multiclass problems

Feature extraction has been an important research topic in pattern classification and has been studied extensively by many researchers. Most of the conventional feature extraction methods are performed using a criterion function defined between two classes or a global function. Although these methods work relatively well in most cases, it is generally not optimal in any sense for multiclass problems. In order to address this problem, the authors propose a method to optimize feature extraction for multiclass problems. The authors first investigate the distribution of classification accuracies of multiclass problems in the feature space and find that there exist much better feature sets that the conventional feature extraction algorithms fail to find. Then the authors propose an algorithm that finds such features. Experiments with remotely sensed data show that the proposed algorithm consistently provides better performances compared with the conventional feature extraction algorithms     

An improved hydrodynamic transport model for silicon

A closed set of hydrodynamic equations for silicon device analysis is obtained with the aid of self-consistent Monte Carlo device simulation data. This set of macroscopic equations is derived without invoking any phenomenological relations such as the Fourier law for heat flow and the Wiedemann-Franz law for thermal conductivity. The model is developed by taking the first four moments of the Boltzmann transport equation (BTE). This model taken into account the difference between the moments of the collision terms of the BTE both for bulk and inhomogeneous systems. The cause of the spurious velocity overshoot sometimes predicted by other models is identified. By introducing different levels of approximation, this system of hydrodynamic equations can be reduced to the conventional hydrodynamic or energy transport equations. The improved model appears to be more accurate than any existing approach for modeling silicon devices     

Mixed pixel classification with robust statistics

The authors present a novel method for mixed pixel classification where the Hough transform and the trimmed means methods are used to classify small sets of pixels. They compare the performance of these methods with the least squares error method, and they show that in the presence of outliers, the trimmed means method is far more reliable than the traditional least squares error method, and even when no outliers are present, its performance is comparable to that of the least squares error method. The method is exhaustively tested using simulated data, and it is also applied to real Landsat TM data for which ground data are available     

A Novel Hierarchical Two-Level Spectral Preconditioning Technique for Electromagnetic Wave Scattering

A new set of higher order hierarchical basis functions based on curvilinear triangular patch is proposed for expansion of the current in electrical field integral equations solved by method of moments. The hierarchical two-level spectral preconditioning technique is developed for the generalized minimal residual iterative method, in which the multilevel fast multipole method is used to accelerate matrix-vector product. The sparse approximate inverse (SAI) preconditioner based on the higher order hierarchical basis functions is used to damp the high frequencies of the error and the low frequencies is eliminated by a spectral preconditioner in a two-level manner defined on the lower order basis functions. The spectral preconditioner is combined with SAI preconditioner to obtain a hierarchical two-level spectral preconditioner. Numerical experiments indicate that the new preconditioner can significantly reduce both the iteration number and computational time.     

Hybrid nonlinear moments in array processing and spectrum analysis

The aim of this paper is to provide a theoretical basis for the use of hybrid nonlinear (HNL) moments in array processing and spectrum analysis. These moments are defined as the expected value of the product of one random variable times a nonlinear function of another random variable. They generalize a class of twofold higher order moments, and their additional flexibility can be exploited for optimization purposes or for computational convenience. A number of properties beyond the classical Bussgang's (1952) and Price's (1958) theorems are found for HNL moments and matrices, making these statistics suitable for harmonic analysis and bearing estimation. Covariance based and higher order moments based methods are extended to the HNL moments domain, and a new class of Gaussian noise rejecting statistics is added to cumulants. The properties of some classes of matrices of HNL moments of practical interest are analyzed in detail     

Microwave scattering model for grass blade structures

The EM scattering solution for a grass blade with complex cross-section geometry is considered. It was previously shown that the scattering solution for such problems, in the form of a polarizability tensor, can be obtained using the low-frequency approximation in conjunction with the method of moments. In addition, that study showed that the relationship between the polarizability tensor of a dielectric cylinder and its dielectric constant can be approximated by a simple algebraic expression. The results of that study are used to show that this algebraic approximation is valid also for cylinders with cross sections the shape of grass blades, providing that proper values are selected for each of three constants appearing in the expression. These constants are dependent on cylinder shape, and if the relationship between the constants and the three parameters describing a grass blade shape can be determined, an algebraic approximation relating polarizability tensor to blade shape, as well as dielectric constant, can be formed. This algebraic approximation can replace the cumbersome method of moments model. The moment method model is therefore used to generate a small but representative set of polarizability tensor data over the range of values commonly observed in nature. A conjugate gradient method is then implemented to correctly determine the three constants of the algebraic approximation for each blade shape. A third-order polynomial fit to the data is then determined for each constant, thus providing a complete analytic replacement to the numerical (moment method) scattering model     

Fuzzy supervised classification of remote sensing images

A fuzzy supervised classification method in which geographical information is represented as fuzzy sets is described. The algorithm consists of two major steps: the estimate of fuzzy parameters from fuzzy training data, and a fuzzy partition of spectral space. Partial membership of pixels allows component cover classes of mixed pixels to be identified and more accurate statistical parameters to be generated, resulting in a higher classification accuracy. Results of classifying a Landsat MSS image are presented, and their accuracy is analyzed