Search

Given a simple and undirected graph, nonnegative node weights, a nonnegative integer j, and a positive integer k, a k- matching in the graph is a subgraph with no isolated [...]

We propose a model to study expectation confirmation in information systems. The proposed model is based on the assimilation-contrast model and prospect theory, and suggests that both are needed to [...]

A measurement-based load updating method for finite element models subjected to static loads was studied using a four-noded curved beam element for large displacements/rotations and a gradient-based variable metric optimizer. Finite element analysis was used to numerically calculate the geometrically linear and nonlinear responses and their sensitivities under a given load. The optimizer was used to recursively update the load so that it minimized the square of the difference between the calculated and prefiltered/noise-free measured (displacement or strain) data. For the basic studies, the extracted load was represented within an element by a linear combination of integrated Legendre polynomials, the coefficients of which were taken as design variables of the least-squares problem. Through a model order analysis, the benefits for solving load updating problems using the relative deformation measurement, the polynomials of lower orders, the elements of larger sizes (because of using the high precision four-noded beam element), and the denser measured points were studied for linear responses under different types of applied loads. It was confirmed that using the reduced number of unknown variables to obtain an overdetermined inverse problem helps get unique and stable extracted loads. Though this conclusion was verified mainly through illustrative examples for a cantilever beam, it was generally applicable to other load updating or inverse problems. Further examples were given for a three-dimensional portal frame load updating, extracting highly oscillating loads, and a strain-based cantilever beam load updating. The final examples were load updating for geometrically nonlinear finite element models under self-weight, snow, and/or pressure loads.

This study reports the validity of the modified Shanks' conjecture on the planar least squares inverse (PLSI) method of stabilizing two-dimensional (2-D) recursive digital filters. A theoretical procedure proposed based on the Lagrange multiplier method of mathematical optimization. The results indicate that the modified Shanks' conjecture reported by Jury was valid for special classes of 2-D polynomials. Key words: PLSI, autocorrelation, filter stability, Lagrange multiplier, INTRODUCTION Two-dimensional (2-D) recursive digital filters find many applications, such as biomedical electronics, image processing, seismic record processing, etc. For a given response characteristic, recursive filters have less hardware requirements [...]

This brief investigates the use of prior knowledge in the parameter estimation of NARMAX polynomial models. The problem of parameter estimation is then formulated in such a way that the estimated models have specified features. This formulation results in a constrained optimization problem, which is solved using the ellipsoid algorithm. This technique is applied to a real dc-dc Buck converter. In this system, the static relation is known from the theory but identification data are located over a rather narrow range around an operating point. Although obtained from dynamical data, the models provide good approximation to the nonlinear static function. Index Terms--Grey-box identification, NARMAX models, parameter estimation, prior knowledge.

The diophantine problem of finding a series of four rational numbers whose products increased by one results in the square of a rational number. The set (1,3,8,120) answers the diophantine problem set. A Eulerian solution was later proposed that proved that are an infinite numbers of four-element sets of positive integers that are diophantine in nature. Variations of the original formula are in use, including a method for finding four-element sets that the product of any three integers is equal to one less than a perfect square.

A pseudo-polynomial dynamic programming algorithm for solving the single-item capacitated economic lot sizing problem with piecewise linear production costs and general holding costs was proposed. Computational experiments showed that the running time of the proposed algorithm is only linearly dependent on the magnitude of the data as demonstrated by its capability to solve quite large problem instances within a reasonable length of time. The average running time of the algorithm grows more or less quadratically when the number of periods are increased while keeping all other characteristics constant. These results proved that the dynamic programming algorithm is a computationally feasible technique with a simple and elegant structure.

The extension of local polynomial fitting with kernel weights to generalized linear and quasi-likelihood conditions is examined. The generalized model assigns an arbitrary but known function of the conditional mean on the response variable. An exponential response distribution for the extension makes the quasi-likelihood function similar to maximum likelihood conditions. Real data applications showed the mathematical advantages of using kernel estimators for polynomial functions., 1. INTRODUCTION Generalized linear models were introduced by Nelder and Wedderburn (1972) as a means of applying techniques used in ordinary linear regression to more general settings. In an important [...]

Doubly graded S(sub n) modules R(sub micro) were described in which it was expected that the multiplicities of irreducible representations in various bihomogenous components are given by the Macdonald coefficients K(sub lambda,micro). The properties and aspects of the construction of the modules R(sub micro) are conjectural, but were proven together with Macdonald positive K. Moreover, it was shown that R(sub micro) can be given several equivalent definitions.

It is well known that kernel regression estimators are subject to so-called boundary or edge effects, a phenomenon in which the bias of an estimator increases near the endpoints of the estimation interval. When the regression curve is linear or nearly linear, the requisite amount of smoothing is so great that the boundary region is effectively the entire estimation interval. Special boundary kernels are proposed here to deal with such cases. It is shown that the proposed kernel estimator has a property also enjoyed by cubic smoothing splines; namely, as the estimator's smoothing parameter becomes large, the estimator tends to a straight line. The limiting straight line is essentially the least squares line when the design points are equally spaced. A simple generalization of ideas in the linear case leads to kernel estimates that are polynomials of any given degree for large bandwidths. Such estimates are an important component of a proposed test for the adequacy of a polynomial model. The test statistic is the bandwidth chosen to minimize an estimated risk function. An example illustrates the usefulness of the new boundary kernels. KEY WORDS: Cross-validation; Goodness of fit; Nonparametric regression; Smoothing parameter selection., 1. INTRODUCTION Kernel smoothers provide a simple and effective method of nonparametric regression estimation. (For an introduction to kernel methods, see Eubank 1988 and Hardle 1990, 1991.) However, for kernel [...]

A quintile rule is derived to approximate the Gini coefficient of income equality, which is exact for fifth-degree polynomials. Examples and solutions are provided for the rule.

Results report a simple polynomial fit function for material gain in the InGaAsP/InP systems. The data on the material-gain model is employed in the semiconductor optical amplfier simulation tools.

The coefficients of the polynomials resulting from the sum of 1(raised to the k power) + 2(raised to the k power) + ... + n(raised to the k power) of the first 'n' positive integers each raised to a nonnegative integer power 'k,' denoted by S(n; k), may be arranged in a triangular array. A proof is presented showing that any row of the triangular array of numbers can be determined by the preceding row. From this proof, an algorithm is developed for obtaining S(n; k+1).

A new method is described to obtain the leading correction to the simple harmonic approximation of a simple pendulum for large-angle motion using an analytical and geometric approach.

'Modeling' data behavior is differentiated from statistically determining a 'best-fit' regression function, as it allows students to develop functions that hold certain attributes in respect to a set of data. Three techniques for developing polynomial models are demonstrated that afford students optional methods for determining appropriate polynomial models for 'curved' data.

The relevant rule of thumb and the polynomials of high degree providing one stays well away from the boundary of the disc of convergence is detailed.

In this note, we show that the zeros of sampled-data systems resulting from rapid sampling of continuous-time systems preceded by a zero-order hold (ZOH) are the roots of the Euler--Frobenius polynomials. Using known properties of these polynomials, we prove two conjectures of Hagiwara and coworkers, the first of which concerns the simplicity, negative realness, and interlacing properties of the sampling zeros of ZOH- and first-order hold (FOH-) sampled systems. To prove the second conjecture, we show that in the fast sampling limit, and as the continuous-time relative degree increases, the largest sampling zero for FOH-sampled systems approaches 1/e, where e is the base of the natural logarithm. Index Terms--Fast sampling, first-order hold, limiting zeros, sampling zeros, zero-order hold.

We discuss two algorithms for mitigating the deleterious effects of nonlinear intersymbol interference in PR4-and EPR4-equalized magnetic recording channels. We call these algorithms RAM-Search algorithms because they borrow the idea of employing random-access memory (RAM) from the RAM decision-feedback equalizer (DFE) and because they involve a search. The RAM-Search algorithms have been devised to mitigate the error propagation problem that arises in making feedforward decisions (for precursor terms) in intersymbol interference (ISI) cancellers used in combination with partial-response signaling. We adopt a simple partial-erasure model for the nonlinear write process and a third-order polynomial model for the nonlinear read process, and show that improvements of about 2.5 dB are possible. We also demonstrate the efficacy of the algorithms on captured spinstand signals that suffer from a significant amount of nonlinear ISI. Index Terms--Magnetic recording channels, partial-erasure model, partial-response channels.

In this paper, the problem of progressive lossless image coding is addressed. A nonlinear decomposition for progressive lossless compression is presented. The decomposition into subbands is called rank-order polynomial decomposition (ROPD) according to the polynomial prediction models used. The decomposition method presented here is a further development and generalization of the morphological subband decomposition (MSD) introduced earlier by the same research group. It is shown that ROPD provides similar or slightly better results than the compared coding schemes such as the codec based on set partitioning in hierarchical trees (SPIHT) and the codec based on wavelet/trellis-coded quantization (WTCQ). Our proposed method highly outperforms the standard JPEG. The proposed lossless compression scheme has the functionality of having a completely embedded bit stream, which allows for data browsing. It is shown that the ROPD has a better lossless rate than the MSD but it has also a much better browsing quality when only a part of the bit stream is decompressed. Finally, the possibility of hybrid lossy/lossless compression is presented using ultrasound images. As with other compression algorithms, considerable gain can be obtained if only the regions of interest are compressed losslessly. Index Terms--Medical image compression, nonlinear subband decomposition, progressive lossless image coding, rank-order polynomial decomposition.

Pattern synthesis of linear antennas utilizing the spherical bessel functions is presented. This leads to antenna current distribution by the Legendre polynomials of the first kind, which are of finite support. Some examples are given to illustrate this procedure. Index Terms--Antenna pattern synthesis.

When estimating a regression function or its derivatives, local polynomials are an attractive choice due to their flexibility and asymptotic performance. Seifert and Gasser proposed ridging of local polynomials to [...]

It is pointed out that main results, including the proof techniques of the recent publication by Curitn and Saba in 2000 and in 1996, are well known and have already appeared in the open literature in Basu and Fetteweis in 1987. The claimed simplification and novelty of proof techniques are not valid because the same method of using the continuity property that the zeros of polynomials have when viewed as functions of coefficients, with adequate refinements, had been used in Basu and Fetteweis in 1987, to arrive at the same set of results as discussed in Curitn and Saba in 2000, and in 1996. Index Terms--Multidimensional systems, recursive digital filters, stability, stable polynomials.

Adjustments are presented to polynomial regression analysis of cohesive soil properties. Topics include various resilience test combinations, such as nonzero strain with zero stress.