Search

Imbalanced domain prediction analysis is currently one of the hot research topics. Many real‐world data mining analyses involve using imbalanced data to obtain predictive models. In the context of imbalance, research on classification problems has been extensive, but research on regression problems is negligible. Rare values rarely occur in imbalanced regression problems, but the focus is on accurately predicting the continuous target variables of rare instances. One of the challenges in imbalanced regression is finding a suitable strategy to rebalance the original dataset in order to improve the predictive performance of the model in rare instances. In this study, two algorithms are proposed: sigma nearest over‐sampling based on convex combination for regression (SNOCCR) and affine combination‐based over‐sampling (ACOS). ACOS rebalances the original dataset by generating new instances through the affine combinations of the original examples. The region where the new instances are generated can be adjusted based on the distribution of the data, ensuring that the generated cases better mimic the distribution of the original examples. The comparison among ACOS, SNOCCR, and other preprocessing methods was conducted on 15 datasets to validate the predictive performance of models trained on rebalanced datasets for rare instances. The experimental results indicate that ACOS outperforms other existing methods. In this study, two over‐sampling algorithms are proposed: sigma nearest over‐sampling based on convex combination for regression (SNOCCR) and affine combination‐based over‐sampling (ACOS) in order to improve the predictive performance of the model in imbalanced regression. The experimental results indicate that ACOS outperforms other existing methods. [ABSTRACT FROM AUTHOR]

Classical symmetric distributions like the Gaussian are widely used. However, in reality data often display a lack of symmetry. Multiple distributions, grouped under the name "skewed distributions", have been developed to specifically cope with asymmetric data. In this paper, we present a broad family of flexible multivariate skewed distributions for which statistical inference is a feasible task. The studied family of multivariate skewed distributions is derived by taking affine combinations of independent univariate distributions. These are members of a flexible family of univariate asymmetric distributions and are an important basis for achieving statistical inference. Besides basic properties of the proposed distributions, also statistical inference based on a maximum likelihood approach is presented. We show that under mild conditions, weak consistency and asymptotic normality of the maximum likelihood estimators hold. These results are supported by a simulation study confirming the developed theoretical results, and some data examples to illustrate practical applicability. [ABSTRACT FROM AUTHOR]

We prove that every quasi-copula can be written as a uniformly converging infinite sum of multiples of copulas. Furthermore, we characterize those quasi-copulas which can be written as a finite sum of multiples of copulas, i.e., that are a linear combination of two copulas. This generalizes a recent result of Fernández-Sánchez, Quesada-Molina, and Úbeda-Flores who considered linear combinations of discrete copulas. [ABSTRACT FROM AUTHOR]

This work presents a data driven control strategy able to track a set point without steady‐state error. The control sequence is computed as an affine combination of past control signals, which belong to a set of trajectories stored in a process historian database. This affine combination is computed so that the variance of the tracking error is minimised. It is shown that offset free control, that is zero mean tracking error, is achieved under the assumption that the state is measurable, the underlying dynamics are linear and the trajectories of the database share the same error dynamics and are in turn offset free. The proposed strategy learns the underlying controller stored in the database while maintaining its offset free tracking capability in spite of differences in the reference, disturbances and operating conditions. No training phase is required and newly obtained process data can be easily taken into account. The proposed strategy, related to direct weight optimisation learning techniques, is tested on a process control trainer. [ABSTRACT FROM AUTHOR]

A mental health trial is analyzed using a dose-response model, in which the number of sessions attended by the patients is deemed indicative of the dose of psychotherapeutic treatment. Here, the parameter of interest is the difference in causal treatment effects between the subpopulations that take part in different numbers of therapy sessions. For this data set, interactions between random treatment allocation and prognostic baseline variables provide the requisite instrumental variables. While the corresponding two-stage least squares (TSLS) estimator tends to have smaller bias than the ordinary least squares (OLS) estimator; the TSLS suffers from larger variance. It is therefore appealing to combine the desirable properties of the OLS and TSLS estimators. Such a trade-off is achieved through an affine combination of these two estimators, using mean squared error as a criterion. This produces the semi-parametric Stein-like (SPSL) estimator as introduced by Judge and Mittelhammer (2004). The SPSL estimator is used in conjunction with multiple imputation with chained equations, to provide an estimator that can exploit all available information. Simulated data are also generated to illustrate the superiority of the SPSL estimator over its OLS and TSLS counterparts. A package entitled SteinIV implementing these methods has been made available through the R platform. © 2017 The Authors. Statistics in Medicine Published by John Wiley & Sons Ltd. [ABSTRACT FROM AUTHOR]

In this paper, a modified Hestenes–Stiefel conjugate gradient method for unconstrained problems is developed, which can achieves the twin goals of generating sufficient descent direction at each iteration as well as being close to the Newton direction. In our methods, the hybridization parameter can also be obtained based on other kinds of conjugacy conditions. Under mild condition, we establish their global convergence for general objective functions. Numerical experimentation with the new method indicates that it efficiently solves the test problems and therefore is promising. [ABSTRACT FROM PUBLISHER]

In this paper, we study and compare different types of generalized barycentric coordinates in detail, including Wachspress, discrete harmonic and mean value coordinates for convex, n-sided polygons. Contour lines are computed in each barycentric coordinate method, and curvature plots of these contour line curves are visualized. Moreover, different distortions of uniform patterns are also shown, providing exact visual method to compare these methods. To overcome the shortcomings of different generalized barycentric computations, affine combination of methods is provided. [ABSTRACT FROM AUTHOR]

• We propose affine combining of two ACLMS filters and present its optimal mixing parameter, based on which an explicit expression describing the steady state mean behavior of the optimal mixing parameter is quantified for doubly white noncircular Gaussian input data. The so established physical insight shows that in steady state, the optimal mixing parameter is, statistically, negative and monotonically increases against the input SO noncircularity coefficient. • Upon applying the optimal mixing parameter, we perform the MSE analysis of the proposed affine-combined ACLMS filters. The so derived explicit MSE expression not only theoretically presents the universality of the combination scheme, that is, smaller or equal steady state MSE than either of the individual ACLMS subfilters, but also demonstrates that the overall MSE in steady state is approximately a monotonic increasing function of the input SO noncircularity coefficient. • Considering that the optimal affine combiner is unacquirable in practical applications, we provide an error power based scheme to iterate the affine combiner. Numerical examples verify that the proposed iterative combiner converges to the optimal one. The overall MSE curves respectively obtained via the iterative and optimal combiners are shown to be in close agreement. This paper investigates the statistical behavior of an affine combination of two augmented complex least mean-square (ACLMS) adaptive filters for processing noncircular Gaussian signals. First, the optimal affine combiner is studied and its steady state mean behavior is explicitly quantified. This makes it possible to reveal the impact of the input second-order noncircularity coefficient on the combiner. Next, the mean square error analysis of the proposed affine-combined ACLMS filters is provided, proving that the combined scheme performs no less than the best of the individual ACLMS subfilters. Besides, to facilitate physically realization of the combination strategy, an error power based algorithm that can iterate the affine combiner to the optimal one is also proposed. Finally, illustrative numerical simulations are provided to support the theoretical results. [ABSTRACT FROM AUTHOR]

Affine combination takes a weighted sum of the outputs of two adaptive filters subjected to the same input whose weighting factor is a parameter calculated using an adaptive mechanism (adaptive filters) in order to achieve a better performance. This study operationalized the behavior of Algorithm Accelerator Version γ (ARγ ) in Affine combination in order to achieve greater efficiency and better performance than the algorithm presented in its basic configuration; likewise identify the effects of the Affine Combination Algorithm Accelerator Version γ (ARγ ) adaptive algorithm. The methodological design and operational plan are part of a research process with a descriptive, explanatory and directed experimental answer why the phenomenon occurs (application of adaptive filters in Affine Combination ) and under what conditions it occurs. [ABSTRACT FROM AUTHOR]

In this paper, we combine the application of different forms of Jensen's inequality to convex and affine combinations. Using this approach we obtained several extensions of the functional form of Jensen's inequality for affine combinations of real valued functions. One of the resulting inequalities is applied to quasi-arithmetic means. [ABSTRACT FROM AUTHOR]

• New second-order interpolative-based M2 maximum-entropy closure for constructing approximate solutions of the radiative transport equation in gray-participating media governed Bose-Einstein statistics. • Review of and new insight into the necessary and sufficient conditions for realizability of predicted moments up to second order in multiple space dimensions. • Demonstration of hyperbolicity of proposed interpolative-based M2 closure via numerical experiments. • Development of boundary conditions for M2 moment closure based on partial moment approach. • Illustration of predictive capabilities of interpolative-based M2 closure for several one and twodimensional canonical test problems. A new interpolative-based approximation to the second-order maximum-entropy, M 2 , moment closure for predicting radiative heat transfer in gray participating media is proposed and described. In addition to preserving many of the desirable mathematical properties of the original M 2 closure, the proposed interpolative approximation provides significant reductions in computational costs compared to the costs of the original M 2 closure by avoiding repeated numerical solution of the corresponding optimization problem for entropy maximization. Theoretical details of the proposed interpolative-based closure, along with a description of an efficient Godunov-type finite-volume scheme that has been developed for the numerical solution of the resulting system of hyperbolic moment equations, are presented. The finite-volume method makes use of limited linear solution reconstruction, multi-block body-fitted quadrilateral meshes with anisotropic adaptive mesh refinement (AMR), and an efficient Newton-Krylov-Schwarz (NKS) iterative method for solution of the resulting non-linear algebraic equations arising from the spatial discretization procedure. The predictive capabilities of the proposed interpolative M 2 closure are assessed by considering a number of model problems involving radiative heat transfer within one- and two-dimensional enclosures, the results for which are compared to solutions of the first-order maximum entropy, M 1 , moment closure, as well as those of the more commonly adopted spherical harmonic moment closure techniques (first-order P 1 and third-order P 3) and the popular discrete ordinates method (DOM). The latter is used as a benchmark for comparisons, whenever exact solutions are not available. The numerical results illustrate the promise of the proposed M 2 closure, with the closure outperforming the M 1 , P 1 and P 3 closures for virtually all cases considered. [ABSTRACT FROM AUTHOR]