The influence of the nonrecent past in prediction for stochastic processes

Consider the stochastic processes X₁, X₂,… and Λ₁, Λ₂,… where the X process can be thought of as observations on the Λ process. We investigate the asymptotic behavior of the conditional distributions of X_t+v given X₁,…, X_t and Λ_t+v given X₁,…, X_t with regard to their dependency on the “early” part of the X process. These distributions arise in various time series and sequential decision theory problems. The results support the intuitively reasonable and often used (as a basic tenet of model building) assumption that only the more recent past is needed for near optimal prediction.     

The contiguity of probability measures and asymptotic inference in continuous time stationary diffusions and Gaussian processes with known covariance

We establish contiguity of families of probability measures indexed by T, as T → ∞, for classes of continuous time stochastic processes which are either stationary diffusions or Gaussian processes with known covariance. In most cases, and in all the examples we consider in Section 4, the covariance is completely determined by observing the process continuously over any finite interval of time. Many important consequences pertaining to properties of tests and estimators, outlined in Section 5, will then apply.     

Bayesian non-parametric estimation for age-dependent branching processes

The purpose of this paper is to obtain Bayes estimators for both the offspring and life-length distribution in the context of a Bellman-Harris age-dependent branching process. We take a non-parametric approach by letting the prior random distributions, for the offspring and life-length distributions, be independent Dirichlet processes. Our primary results concern the derivation of Bayes estimators, under weighted squared error loss for each distribution. We also indicate some of their asymptotic properties and briefly discuss the modifications that become necessary when the initial information is such that the prior random distribution cannot be taken to be independent.     

On Nummelin splitting for continuous time Harris recurrent Markov processes and application to kernel estimation for multi-dimensional diffusions

We introduce a sequence of stopping times that allow us to study an analogue of a life-cycle decomposition for a continuous time Markov process, which is an extension of the well-known splitting technique of Nummelin to the continuous time case. As a consequence, we are able to give deterministic equivalents of additive functionals of the process and to state a generalisation of Chen’s inequality. We apply our results to the problem of non-parametric kernel estimation of the drift of multi-dimensional recurrent, but not necessarily ergodic, diffusion processes.     

Limit theorems for stochastic measures of the accuracy of density estimators

Stochastic measures of the distance between a density f and its estimate f_n have been used to compare the accuracy of density estimators in Monte Carlo trials. The practice in the past has been to select a measure largely on the basis of its ease of computation, using only heuristic arguments to explain the large sample behaviour of the measure. Steele [11] has shown that these arguments can lead to incorrect conclusions. In the present paper we obtain limit theorems for the stochastic processes derived from stochastic measures, thereby explaining the large sample behaviour of the measures.     

On the local asymptotic behavior of the likelihood function for Meixner Lévy processes under high-frequency sampling

We discuss the local asymptotic behavior of the likelihood function associated with all the four characterizing parameters (α,β,δ,μ) of the Meixner Lévy process under high-frequency sampling scheme. We derive the optimal rate of convergence for each parameter and the Fisher information matrix in a closed form. The skewness parameter β exhibits a slower rate alone, relative to the other three parameters free of sampling rate. An unusual aspect is that the Fisher information matrix is constantly singular for full joint estimation of the four parameters. This is a particular phenomenon in the regular high-frequency sampling setting and is of essentially different nature from low-frequency sampling. As soon as either α or δ is fixed, the Fisher information matrix becomes diagonal, implying that the corresponding maximum likelihood estimators are asymptotically orthogonal.     

The asymptotic distribution of the likelihood ratio for autoregressive time series with a regression trend

It is shown that the likelihood ratio of an autoregressive time series of finite order with a regression trend is asymptotically normal. This result is used to derive the power of a test for positive correlation of the residuals under local autoregressive alternatives. The test is based on the Durbin-Watson statistics.     

Nonnull distribution of likelihood ratio criterion for reality of covariance matrix

In this paper the distribution of the likelihood ratio test for testing the reality of the covariance matrix of a complex multivariate normal distribution is investigated. Some simplifications in the noncentral distribution are made and the noncentral distribution is derived for the special case where the rank of the noncentrality matrix is two. In the null case exact expressions for the distribution are given up to p = 6, and percentage points are tabulated. These percentage points were compared with percentage points derived from an asymptotic expansion of the distribution, and the accuracy of the approximation was found to be sufficient for several practical situations.     

Stochastic processes with independent increments,taking values in a Hilbert space

Let $\text{H}$ be a real separable Hilbert space; let X(t), t?[0, 1], be a separable, stochastically continuous, $\text{H}$-valued stochastic process with independent increments. Then a decomposition of X(t) into a uniformly convergent sum of independent processes is found. In this decomposition one of the processes is Gaussian with continuous sample functions, and the remainder of the processes have sample functions whose discontinuities correspond to those of certain real-valued Poisson processes. The decomposition of X(t) leads to a Levy-Khintchine representation of the characteristic functional of X(t). In addition, the case when X(t) has finite variance is explored, and, as a consequence of the above decomposition, a Kolmogorov-type representation of the characteristic functional of X(t) is derived.     

Two strong limit theorems for processes with independent increments

Two related almost sure limit theorems are obtained in connection with a stochastic process {ξ(t), ?∞ < t < ∞} with independent increments. The first result deals with the existence of a simultaneous stabilizing function H(t) such that $\text{(ξ(t) ? ξ(0))}\text{H(t)}\text{→ 0}$ for almost all sample functions of the process. The second result deals with a wide-sense stationary process whose random spectral distributions is ξ. It addresses the question: Under what conditions does $\text{(2T)}{}^{\mathrm{?1}}\text{∫}{}_{\mathrm{?T}}{}^{\mathrm{T}}\text{X(t)}\text{X(t + τ)}\text{dt}$ converge as T → ∞ for all τ for almost all sample functions?     

The likelihood ratio and its applications in sequential analysis

The likelihood ratio is treated as a generalized Radon-Nikodym derivative for measures which need not be absolutely continuous. It is shown that a wide variety of identities and inequalities in sequential analysis follow directly from simple properties of likelihood ratios.     

A conditional dichotomy theorem for stochastic processes with independent increments

Let X(t) and Y(t) be two stochastically continuous processes with independent increments over [0, T] and Lévy spectral measures M_t and N_t, respectively, and let the “time-jump” measures M and N be defined over [0, T] × $\text{R}$?{0} by M((t₁, t₂] × A) = M_t2(A) ? M_t1(A) and N((T₁, t₂] × A) = N_t2(A) ? N_t1(A). Under the assumption that M is equivalent to N, it is shown that the measures induced on function space by X(t) and Y(t) are either equivalent or orthogonal, and necessary and sufficient conditions for equivalence are given. As a corollary a complete characterization of the set of admissible translates of such processes is obtained: a function f is an admissible translate for X(t) if and only if it is an admissible translate for the Gaussian component of X(t). In particular, if X(t) has no Gaussian component, then every nontrivial translate of X(t) is orthogonal to it.     

On processes of ornstein-uhlenbeck type in hilbert space

A process fo Ornstein-Uhlenbeck type is a mild solution of the stochastic differential system in Hilbert space dXt=AX _t dt+dZ _t, where A generates a semigroup of operators and Z _tis a process with homogeneous independent increments. The explicit integral formula for the process of O-U type is given. The main purpose is to study stationary distributions for such processes. Sufficient and necessary conditions for existence and characterization are given. The difference between finite and infinite dimensional cases is illustrated by examples     

On exponential rates of likelihood ratio estimators for location parameters

In this paper the exponential rates, bounds, and local exponential rates for likelihood ratio estimators are studied. Under certain regularity conditions, a family of likelihood ratio estimators is shown to be admissible in exponential rate. It is also shown that the maximum likelihood estimator is the limit of this family of estimators.     

Multivariate measures of skewness for the skew-normal distribution

The main objective of this work is to calculate and compare different measures of multivariate skewness for the skew-normal family of distributions. For this purpose, we consider the Mardia (1970) [10], Malkovich and Afifi (1973) [9], Isogai (1982) [17], Srivastava (1984) [15], Song (2001) [14], Móri et al. (1993) [11], Balakrishnan et al. (2007) [3] and Kollo (2008) [7] measures of skewness. The exact expressions of all measures of skewness, except for Song’s, are derived for the family of skew-normal distributions, while Song’s measure of shape is approximated by the use of delta method. The behavior of these measures, their similarities and differences, possible interpretations, and their practical use in testing for multivariate normal are studied by evaluating their power in the case of some specific members of the multivariate skew-normal family of distributions.     

Sampling distribution for a class of estimators for nonregular linear processes

Let {X_t; t = 1, 2,…} be a linear process with a location parameter θ defined by X_t ? θ = Σ₀^∞g_rZ_t?r where {Z_t; t = 0, ±1,…} is a sequence of independent and identically distributed random variables, with E∥Z₁∥^δ < ∞ for some δ > 0. If δ ? 1 we assume further than E(Z₁) = 0. Let η = δ if 0 < δ < 2, and η = 2 if δ ? 2. Then assume that Σ₀^∞∥ g_r ∥^η < ∞. Consider the class of estimators $\text{θ}\text{∧}{}_{\mathrm{n}}$ given by $\text{θ}\text{∧}{}_{\mathrm{n}}\text{= Σ}{}_1{}^{\mathrm{n}}\text{c}{}_{\mathrm{nt}}\text{X}{}_{\mathrm{t}}\text{where}\text{c}{}_{\mathrm{nt}}$ is of the form c_nt = Σ_{p = 0}^sβ_npt^p for some s ? 0. An attempt has been made to investigate the distributional properties of $\text{θ}\text{∧}{}_{\mathrm{n}}$ in large samples for various choices of β_np (0 ? p ? s), s, and the distribution of Z₁ under the constraints Σ₀^∞r^kg_r = 0, 0 ? k ? q where q in an arbitrary integer, 0 ? q ? s.     

On estimator efficiency in stochastic processes

It is shown, under mild regularity conditions on the random information matrix, that the maximum likelihood estimator is efficient in the sense of having asymptotically maximum probability of concentration about the true parameter value. In the case of a single parameter, the conditions are improvements of those used by Heyde (1978). The proof is based on the idea of maximum probability estimators introduced by Weiss and Wolfowitz (1967).