Maximum likelihood estimation in the presence of outiliers

This paper deals with the maximum likelihood estimation of parameters when the sample (x₁…x_n ) may heve k spuriously generated observations from another distribution, say G≠F, where F is the distribution of the target population. If G is stochastically larger than F, then these k observations may give rise to k extreme observations or ‘outliers’. This situation is often described by a so-called ‘k-outlier model’ in which in addition to the parameters involved in F and G, the set ν={ν₁,…,ν_k} of indices, for which x_νj , j=1,…,k, come from G, is also unknow.     

Approximate maximum likelihood estimation in the presence of extra-poisson variation

A two-stage hierarchical model for analysis of discrete data with extra-Poisson variation is examined. The model consists of a Poisson distribution with a mixing lognormal distribution for the mean. A method of approximate maximum likelihood estimation of the parameters is proposed. The method uses the EM algorithm and approximations to facilitate its implementation are derived. Approximate standard errors of the estimates are provided and a numerical example is used to illustrate the method.     

Quasi-maximum likelihood estimation of GARCH models in the presence of missing values

This work presents a new method to deal with missing values in financial time series. Previous works are generally based in state-space models and Kalman filter and few consider ARCH family models. The traditional approach is to bound the data together and perform the estimation without considering the presence of missing values. The existing methods generally consider missing values in the returns. The proposed method considers the presence of missing values in the price of the assets instead of in the returns. The performance of the method in estimating the parameters and the volatilities is evaluated through a Monte Carlo simulation. Value at risk is also considered in the simulation. An empirical application to NASDAQ 100 Index series is presented.     

Maximum likelihood estimation for arma models in the presence of ARMA errors

An ARMA(p, q) process observed with an ARMA(c, d) error has an ARMA (p + c, k) representation with k = max(c + q, p + d) whose parameters satisfy some nonlinear constraints. Identification of the model is discussed. We develop Newton-Raphson estimators for the ARMA(p + c, k) process which take advantage of the information contained in the nonlinear restrictions. Explicit expressions for the derivatives of the restrictions are derived.     

A new algorithm for maximum likelihood estimation in normal scale-mixture generalized autoregressive conditional heteroskedastic models

In this paper, we propose a new generalized autoregressive conditional heteroskedastic (GARCH) model using infinite normal scale-mixtures which can suitably avoid order selection problems in the application of finite normal scale-mixtures. We discuss its theoretical properties and develop a two-stage algorithm for the maximum likelihood estimator to estimate the mixing distribution non-parametric maximum likelihood estimator (NPMLE) as well as GARCH parameters (two-stage MLE). For the estimation of a mixing distribution, we employ a fast computational algorithm proposed by Wang [On fast computation of the non-parametric maximum likelihood estimate of a mixing distribution. J R Stat Soc Ser B. 2007;69:185–198] under the gradient characterization of the non-parametric mixture likelihood. The GARCH parameters are then estimated either using the expectation-mazimization algorithm or general optimization scheme. In addition, we propose a new forecasting algorithm of value-at-risk (VaR) using the two-stage MLE and the NPMLE. Through a simulation study and real data analysis, we compare the performance of the two-stage MLE with the existing ones including quasi-maximum likelihood estimator based on the standard normal density and the finite normal mixture quasi maximum estimated-likelihood estimator (cf. Lee S, Lee T. Inference for Box–Cox transformed threshold GARCH models with nuisance parameters. Scand J Stat. 2012;39:568–589) in terms of the relative efficiency and accuracy of VaR forecasting.     

On asymptotic normality of likelihood and conditional analysis

The likelihood function from a large sample is commonly assumed to be approximately a normal density function. The literature supports, under mild conditions, an approximate normal shape about the maximum; but typically a stronger result is needed: that the normalized likelihood itself is approximately a normal density. In a transformation-parameter context, we consider the likelihood normalized relative to right-invariant measure, and in the location case under moderate conditions show that the standardized version converges almost surely to the standard normal. Also in a transformation-parameter context, we show that almost sure convergence of the normalized and standardized likelihood to a standard normal implies that the standardized distribution for conditional inference converges almost surely to a corresponding standard normal. This latter result is of immediate use for a range of estimating, testing, and confidence procedures on a conditional-inference basis.     

On the finite-sample properties of conditional empirical likelihood estimators

We provide Monte Carlo evidence on the finite-sample behavior of the conditional empirical likelihood (CEL) estimator of Kitamura, Tripathi, and Ahn and the conditional Euclidean empirical likelihood (CEEL) estimator of Antoine, Bonnal, and Renault in the context of a heteroscedastic linear model with an endogenous regressor. We compare these estimators with three heteroscedasticity-consistent instrument-based estimators and the Donald, Imbens, and Newey estimator in terms of various performance measures. Our results suggest that the CEL and CEEL with fixed bandwidths may suffer from the no-moment problem, similarly to the unconditional generalized empirical likelihood estimators studied by Guggenberger. We also study the CEL and CEEL estimators with automatic bandwidths selected through cross-validation. We do not find evidence that these suffer from the no-moment problem. When the instruments are weak, we find CEL and CEEL to have finite-sample properties—in terms of mean squared error and coverage probability of confidence intervals—poorer than the heteroscedasticity-consistent Fuller (HFUL) estimator. In the strong instruments case, the CEL and CEEL estimators with automatic bandwidths tend to outperform HFUL in terms of mean squared error, while the reverse holds in terms of the coverage probability, although the differences in numerical performance are rather small.     

Maximum likelihood estimation for bivariate SUR Tobit modeling in presence of two right-censored dependent variables

This article extends the analysis of the Seemingly Unrelated Regression (SUR) Tobit model for two right-censored dependent variables by modeling its nonlinear dependence structure through the rotated by 180 degrees version of the Clayton copula. An advantage of our approach is to provide unbiased point estimates of the marginal and copula parameters. Moreover, we discuss the construction of confidence intervals using bootstrap resampling procedures. The results of the performed simulation study demonstrate the good performance of the proposed methods. We illustrate our procedures using bivariate customer churn data from a Brazilian commercial bank.     

Non-parametric estimation of the conditional mode

The problem of estimating the mode of a conditional probability density function is considered. It is shown that under some regularity conditions the estimate of the conditional mode obtained by maximizing a kernel estimate of the conditional probability density function is strongly consistent and asymptotically normally distributed.     

Further remarks on asymptotic normality of likelihood and conditional analyses

Under weak conditions the normalized likelihood with or without weight function almost surely converges to a normal density function: for a real parameter or a vector parameter; with or without the assumption of independent identical distributions. Applications arise for confidence intervals, confidence distributions, structural distributions. and conditional analyses with transformation and structural models.     

Empirical likelihood inference in the presence of measurement error

Suppose that several different imperfect instruments and one perfect instrument are used independently to measure some characteristic of a population. The authors consider the problem of combining this information to make statistical inference on parameters of interest, in particular the population mean and cumulative distribution function. They develop maximum empirical likelihood estimators and study their asymptotic properties. They also present simulation results on the finite sample efficiency of these estimators.     

Asymptotic properties of the MAMSE adaptive likelihood weights

The weighted likelihood is a generalization of the likelihood designed to borrow strength from similar populations while making minimal assumptions. If the weights are properly chosen, the maximum weighted likelihood estimate may perform better than the maximum likelihood estimate (MLE). In a previous article, the minimum averaged mean squared error (MAMSE) weights are proposed and simulations show that they allow to outperform the MLE in many cases. In this paper, we study the asymptotic properties of the MAMSE weights. In particular, we prove that the MAMSE-weighted mixture of empirical distribution functions converges uniformly to the target distribution and that the maximum weighted likelihood estimate is strongly consistent. A short simulation illustrates the use of bootstrap in this context.     

Maximum likelihood estimation for the beta distribution

A fast method of calculating the two-parameter maximum-likelihood estimates of the beta distribution is given which does not require starting values and is generally free from convergence problems.     

Sequential change detection in the presence of unknown parameters

It is commonly required to detect change points in sequences of random variables. In the most difficult setting of this problem, change detection must be performed sequentially with new observations being constantly received over time. Further, the parameters of both the pre- and post- change distributions may be unknown. In Hawkins and Zamba (Technometrics 47(2):164–173, 2005), the sequential generalised likelihood ratio test was introduced for detecting changes in this context, under the assumption that the observations follow a Gaussian distribution. However, we show that the asymptotic approximation used in their test statistic leads to it being conservative even when a large numbers of observations is available. We propose an improved procedure which is more efficient, in the sense of detecting changes faster, in all situations. We also show that similar issues arise in other parametric change detection contexts, which we illustrate by introducing a novel monitoring procedure for sequences of Exponentially distributed random variable, which is an important topic in time-to-failure modelling.     

Maximum likelihood estimation of the logarithmic series distribution

This paper discusses the maximum likelihood estimation of the parameter of the logarithmic series distribution. The univariate case is treated in Part I, the multivariate case in Part II. A simple numerical estimation procedure is suggested using a fixed point approach. Convergence to the maximum likelihood estimator is shown. In Part III convergence rate is proven to be linear which is also demonstrated through example. In addition, comparisons with Newton’s method and the secant method in the univariate case, and with Newton’s method and the projected gradient method in the multivariate case are provided.     

A linear model-based test for the heterogeneity of conditional correlations

Current methods of testing the equality of conditional correlations of bivariate data on a third variable of interest (covariate) are limited due to discretizing of the covariate when it is continuous. In this study, we propose a linear model approach for estimation and hypothesis testing of the Pearson correlation coefficient, where the correlation itself can be modeled as a function of continuous covariates. The restricted maximum likelihood method is applied for parameter estimation, and the corrected likelihood ratio test is performed for hypothesis testing. This approach allows for flexible and robust inference and prediction of the conditional correlations based on the linear model. Simulation studies show that the proposed method is statistically more powerful and more flexible in accommodating complex covariate patterns than the existing methods. In addition, we illustrate the approach by analyzing the correlation between the physical component summary and the mental component summary of the MOS SF-36 form across a fair number of covariates in the national survey data.     

Composite likelihood estimation in multivariate data analysis

The authors propose two composite likelihood estimation procedures for multivariate models with regression/univariate and dependence parameters. One is a two‐stage method based on both univariate and bivariate margins. The other estimates all the parameters simultaneously based on bivariate margins. For some special cases, the authors compare their asymptotic efficiencies with the maximum likelihood method. The performance of the two methods is reasonable, except that the first procedure is inefficient for the regression parameters under strong dependence. The second approach is generally better for the regression parameters, but less efficient for the dependence parameters under weak dependence.     

Maximum likelihood estimation in convex hull models

In this paper we consider models involving the convex hull operation of the parameter and the noise i.e. Y_i = CH(A, X_X). Then we generalize the basic models to ANOVA models; i.e. Y_ij=CH(A∪B_j,X_ij). In some cases the consistent estimators for the J U new parameters are derived. Assuming the existence of density forrandom convex sets, we derive the likelihood for the convex hull model. We then find the maximum Likelihood Estimators for the parameters. Examples for some random convex sets with finite dimensional distributions are derived to show how good these estimators are.     

On maximum likelihood estimation of the binomial parameter n

This paper gives a necessary and sufficient condition for the existence of a finite conditional maximum-likelihood estimate for the binomial parameter n. Upper bounds for the conditional and beta-binomial maximum-likelihood estimators are derived. An example is given to show that the conditional likelihood function and the beta-binomial likelihood function may not be unimodal.