Testing in linear composite quantile regression models

Composite quantile regression (CQR) can be more efficient and sometimes arbitrarily more efficient than least squares for non-normal random errors, and almost as efficient for normal random errors. Based on CQR, we propose a test method to deal with the testing problem of the parameter in the linear regression models. The critical values of the test statistic can be obtained by the random weighting method without estimating the nuisance parameters. A distinguished feature of the proposed method is that the approximation is valid even the null hypothesis is not true and power evaluation is possible under the local alternatives. Extensive simulations are reported, showing that the proposed method works well in practical settings. The proposed methods are also applied to a data set from a walking behavior survey.     

A generalized Newton algorithm for quantile regression models

This paper formulates the quadratic penalty function for the dual problem of the linear programming associated with the \(L_1\) constrained linear quantile regression model. We prove that the solution of the original linear programming can be obtained by minimizing the quadratic penalty function, with the formulas derived. The obtained quadratic penalty function has no constraint, thus could be minimized efficiently by a generalized Newton algorithm with Armijo step size. The resulting algorithm is easy to implement, without requiring any sophisticated optimization package other than a linear equation solver. The proposed approach can be generalized to the quantile regression model in reproducing kernel Hilbert space with slight modification. Extensive experiments on simulated data and real-world data show that, the proposed Newton quantile regression algorithms can achieve performance comparable to state-of-the-art.     

Composite quantile regression for single-index models with asymmetric errors

For the single-index model, a composite quantile regression technique is proposed in this paper to construct robust and efficient estimation. Theoretical analysis reveals that the proposed estimate of the single-index vector is highly efficient relative to its corresponding least squares estimate. For the single-index vector, the proposed method is always valid across a wide spectrum of error distributions; even in the worst case scenario, the asymptotic relative efficiency has a lower bound 86.4 %. Meanwhile, we employ weighted local composite quantile regression to obtain a consistent and robust estimate for the nonparametric component in the single-index model, which is adapted to both symmetric and asymmetric distributions. Numerical study and a real data analysis can further illustrate our theoretical findings.     

Markov regime-switching quantile regression models and financial contagion detection

In this paper, we propose a Markov regime-switching quantile regression model, which considers the case where there may exist equilibria jumps in quantile regression. The parameters are estimated by the maximum likelihood estimation (MLE) method. A simulation study of this new model is conducted covering many scenarios. The simulation results show that the MLE method is efficient in estimating the model parameters. An empirical analysis is also provided, which focuses on the detection of financial crisis contagion between United States and some European Union countries during the period of sub-prime crisis from the angle of financial risk. The degree of financial contagion between markets is subsequently measured by utilizing the quantile regression coefficients. The empirical results show that in a crisis situation, the interdependence between United States and European Union countries dramatically increases.     

Polynomial network autoregressive models with divergent orders

We propose a novel polynomial network autoregressive model by incorporating higher-order connected relationships to simultaneously model the effects of both direct and indirect connections. A quasimaximum likelihood estimation method is proposed to estimate the unknown influence parameters, and we demonstrate its consistency and asymptotic normality without imposing any distribution assumption. Moreover,an extended Bayesian information criterion is set for order selection with a divergent upper ...     

Polynomial diffusion models for life insurance liabilities

In this paper we study the pricing and hedging problem of a portfolio of life insurance products under the benchmark approach, where the reference market is modelled as driven by a state variable following a polynomial diffusion on a compact state space. Such a model can be used to guarantee not only the positivity of the OIS short rate and the mortality intensity, but also the possibility of approximating both pricing formula and hedging strategy of a large class of life insurance products by explicit formulas.     

Confidence bands for quantile function under random censorship

Some new confidence bands are established for the quantile function from randomly censored data. The method does not require estimation of the density function. As an application, we construct bands for the quantile function of the length of fractures in the granitic plutons near Lac du Bonnet, Manitoba, where an Underground Research Laboratory is being built for the nuclear waste disposal program in Canada.Geological Survey of Canada Contribution Number 10386.Research partially supported by an NSERC Canada Grant.Research done while at Carieton University, supported by NSERC Canada and EMR Canada Grants of M. Csörg.     

Marginal quantile regression for varying coefficient models with longitudinal data

In this paper, we investigate the quantile varying coefficient model for longitudinal data, where the unknown nonparametric functions are approximated by polynomial splines and the estimators are obtained by minimizing the quadratic inference function. The theoretical properties of the resulting estimators are established, and they achieve the optimal convergence rate for the nonparametric functions. Since the objective function is non-smooth, an estimation procedure is proposed that uses induced smoothing and we prove that the smoothed estimator is asymptotically equivalent to the original estimator. Moreover, we propose a variable selection procedure based on the regularization method, which can simultaneously estimate and select important nonparametric components and has the asymptotic oracle property. Extensive simulations and a real data analysis show the usefulness of the proposed method.

     

Semi-parametric quantile estimation for double threshold autoregressive models with heteroskedasticity

Compared to the conditional mean or median, conditional quantiles provide a more comprehensive picture of a variable in various scenarios. A semi-parametric quantile estimation method for a double threshold auto-regression with exogenous regressors and heteroskedasticity is considered, allowing representation of both asymmetry and volatility clustering. As such, GARCH dynamics with nonlinearity are added to a nonlinear time series regression model. An adaptive Bayesian Markov chain Monte Carlo scheme, exploiting the link between the quantile loss function and the asymmetric-Laplace distribution, is employed for estimation and inference, simultaneously estimating and accounting for nonlinear heteroskedasticity plus unknown threshold limits and delay lags. A simulation study illustrates sampling properties of the method. Two data sets are considered in the empirical applications: modelling daily maximum temperatures in Melbourne, Australia; and exploring dynamic linkages between financial markets in the US and Hong Kong.     

Polynomial viscosity methods for multispecies kinematic flow models

Multispecies kinematic flow models are defined by systems of strongly coupled, nonlinear first‐order conservation laws. They arise in various applications including sedimentation of polydisperse suspensions and multiclass vehicular traffic. Their numerical approximation is a challenge since the eigenvalues and eigenvectors of the corresponding flux Jacobian matrix have no closed algebraic form. It is demonstrated that a recently introduced class of fast first‐order finite volume solvers, called polynomial viscosity matrix (PVM) methods [M. J. Castro Díaz and E. Fernández‐Nieto, SIAM J Sci Comput 34 (2012), A2173–A2196], can be adapted to multispecies kinematic flows. PVM methods have the advantage that they only need some information about the eigenvalues of the flux Jacobian, and no spectral decomposition of a Roe matrix is needed. In fact, the so‐called interlacing property (of eigenvalues with known velocity functions), which holds for several important multispecies kinematic flow models, provides sufficient information for the implementation of PVM methods. Several variants of PVM methods (differing in polynomial degree and the underlying quadrature formula to approximate the Roe matrix) are compared by numerical experiments. It turns out that PVM methods are competitive in accuracy and efficiency with several existing methods, including the Harten, Lax, and van Leer method and a spectral weighted essentially non‐oscillatory scheme that is based on the same interlacing property. © 2016 Wiley Periodicals, Inc. Numer Methods Partial Differential Eq 32: 1265–1288, 2016     

Ageing concepts: An approach based on quantile function

The notion of ageing plays an important role in reliability analysis and in identifying life distributions. Most of the ageing concepts existing in the literature are described on the basis of measures defined in terms of the distribution function. Recently, the role of quantile functions has also been identified as lifetime models, and reliability functions based on distribution functions were redefined in terms of quantile functions. In the present paper, we redefine some important popular ageing concepts using quantile functions. The uses of new definitions are illustrated by discussing ageing properties of some quantile function models.     

Likelihood ratio-type tests in weighted composite quantile regression of DTARCH models

Science China Mathematics - The double-threshold autoregressive conditional heteroscedastic (DTARCH) model is a useful tool to measure and forecast the mean and volatility of an asset return in a...     

Polynomial combinatorial algorithms for skew-bisubmodular function minimization

Mathematical Programming - Huber et al. (SIAM J Comput 43:1064–1084, 2014) introduced a concept of skew bisubmodularity, as a generalization of bisubmodularity, in their complexity dichotomy...     

Polynomial spline estimation for partial functional linear regression models

Because of its orthogonality, interpretability and best representation, functional principal component analysis approach has been extensively used to estimate the slope function in the functional linear model. However, as a very popular smooth technique in nonparametric/semiparametric regression, polynomial spline method has received little attention in the functional data case. In this paper, we propose the polynomial spline method to estimate a partial functional linear model. Some asymptotic results are established, including asymptotic normality for the parameter vector and the global rate of convergence for the slope function. Finally, we evaluate the performance of our estimation method by some simulation studies.     

The cumulative quantile regression function with censored and truncated covariate

In this paper, we consider the empirical estimator of the cumulative quantile regression (CQR) functionwhen the covariate is subjected to random truncation and censorship. Strong Gaussian approximations for the associated CQR process are established under appropriate assumptions. A functional law of the iterated logarithm for the CQR process is also derived. These results provide a foundation for the asymptotic theory of functional statistics based on these processes.     

Upper and lower approximation models in interval regression using regression quantile techniques

In this paper, we propose new interval regression analysis based on the regression quantile techniques. To analyze a phenomenon in a fuzzy environment, we propose two interval approximation models. Without using all data, we first identify the main trend from the designated proportion of the given data. To select the main part of data to be analyzed, we introduce the regression quantile techniques. The obtained model is not influenced by extreme points since it is formulated from the center-located main proportion of the given data. After that, the interval regression model including all data can be identified based on the acquired main trend. The obtained interval regression model by the main proportion of the given data is called the lower approximation model, while interval regression model by all data is called the upper approximation model for the given phenomenon. Also it is shown that, from the lower approximation model (main trend) and the upper approximation model, we can construct a trapezoidal fuzzy model. The membership function of this fuzzy model is useful to obtain the locational information for each observation. The characteristic of our approach can be described as obtaining the upper and lower approximation models and combining them to be a fuzzy model for representing the given phenomenon in a fuzzy environment.     

部分线性单指标模型的复合分位数回归及变量选择

本文提出复合最小化平均分位数损失估计方法 (composite minimizing average check loss estimation,CMACLE)用于实现部分线性单指标模型(partial linear single-index models,PLSIM)的复合分位数回归(composite quantile regression,CQR).首先基于高维核函数构造参数部分的复合分位数回归意义下的相合估计,在此相合估计的基础上,通过采用指标核函数进一步得到参数和非参数函数的可达最优收敛速度的估计,并建立所得估计的渐近正态性,比较PLSIM的CQR估计和最小平均方差估计(MAVE)的相对渐近效率.进一步地,本文提出CQR框架下PLSIM的变量选择方法,证明所提变量选择方法的oracle性质.随机模拟和实例分析验证了所提方法在有限样本时的表现,证实了所提方法的优良性.     

Polynomial function and derivative approximation of Sinc data

Sinc methods consist of a family of one dimensional approximation procedures for approximating nearly every operation of calculus. These approximation procedures are obtainable via operations on Sinc interpolation formulas. Nearly all of these approximations–except that of differentiation–yield exceptional accuracy. The exception: when differentiating a Sinc interpolation formula that gives an approximation over an interval with a finite end-point. In such cases, we obtain poor accuracy in the neighborhood of the finite end-point. In this paper we derive novel polynomial-like procedures for differentiating a function that is known at Sinc points, to obtain an approximation of the derivative of the function that is uniformly accurate on the whole interval, finite or infinite, in the case when the function itself has a derivative on the closed interval.     

Estimation of a smooth quantile function under the proportional hazards model

The problem of estimating a smooth quantile function, Q(·), at a fixed point p, 0 < p < 1, is treated under a nonparametric smoothness condition on Q. The asymptotic relative deficiency of the sample quantile based on the maximum likelihood estimate of the survival function under the proportional hazards model with respect to kernel type estimators of the quantile is evaluated. The comparison is based on the mean square errors of the estimators. It is shown that the relative deficiency tends to infinity as the sample size, n, tends to infinity.     

Performance of balanced bootstrap resampling in distribution function and quantile problems

Summary It is shown that balanced resampling improves on ordinary uniform resampling when employed to estimate distribution functions or quantiles, but that the improvement is not by an order of magnitude as happens in problems such as bias estimation. Maximum improvement occurs at the centre of the distribution, where performance increases by 175%. The increase is only 16% at the 2.5% and 97.5% quantiles. In general the improvement is by the factor {1–²(1–)}^–1, where is the standard normal distribution function and =.