Uniform Accuracy of the Maximum Likelihood Estimates for Probabilistic Models of Biological Sequences

Probabilistic models for biological sequences (DNA and proteins) have many useful applications in bioinformatics. Normally, the values of parameters of these models have to be estimated from empirical data. However, even for the most common estimates, the maximum likelihood (ML) estimates, properties have not been completely explored. Here we assess the uniform accuracy of the ML estimates for models of several types: the independence model, the Markov chain and the hidden Markov model (HMM). Particularly, we derive rates of decay of the maximum estimation error by employing the measure concentration as well as the Gaussian approximation, and compare these rates.     

一致椭圆扩散过程的几类极大游程

本文研究了暂留一致椭圆扩散过程，利用强Markov性，求出了在首中此球之前、末离此球之前和在首中此球与末离此球之间过程的几类极大游程的分布的估计，推广了文献、相关的结果．     

Assessing Convergence and Mixing of MCMC Implementations via Stratification

Some posterior distributions lead to Markov chain Monte Carlo (MCMC) chains that are naturally viewed as collections of subchains. Examples include mixture models, regime-switching models, and hidden Markov models. We obtain MCMC-based estimators of posterior expectations by combining different subgroup (subchain) estimators using stratification and poststratification methods. Variance estimates of the limiting distributions of such estimators are developed. Based on these variance estimates, we propose a test statistic to aid in the assessment of convergence and mixing of chains. We compare our diagnostic with other commonly used methods. The approach is illustrated in two examples: a latent variable model for arsenic concentration in public water systems in Arizona and a Bayesian hierarchical model for Pacific sea surface temperatures. Supplementary materials, which include MATLAB codes for the proposed method, are available online.     

Examples of fitting structured phase–type distributions

A sub–class of phase–type distributions is defined in terms of a Markov process with sequential transitions between transient states and transitions from these states to absorption. Such distributions form a very rich class; they can be fitted to data, and any structure revealed by the parameter estimates used to develop more parsimonious re–parametrizations. Several example data sets are used as illustrations.     

Stochastic processes with orthogonal polynomial eigenfunctions

Markov processes which are reversible with either Gamma, Normal, Poisson or Negative Binomial stationary distributions in the Meixner class and have orthogonal polynomial eigenfunctions are characterized as being processes subordinated to well-known diffusion processes for the Gamma and Normal, and birth and death processes for the Poisson and Negative Binomial. A characterization of Markov processes with Beta stationary distributions and Jacobi polynomial eigenvalues is also discussed.     

An invariance principle for non-symmetric Markov processes and reflecting diffusions in random domains

Summary We study an invariance principle for additive functionals of nonsymmetric Markov processes with singular mean forward velocities. We generalize results of Kipnis and Varadhan [KV] and De Masi et al. [De] in two directions: Markov processes are non-symmetric, and mean forward velocities are distributions. We study continuous time Markov processes. We use our result to homogenize non-symmetric reflecting diffusions in random domains.     

Heteroscedastic nonlinear regression models based on scale mixtures of skew-normal distributions

An extension of some standard likelihood based procedures to heteroscedastic nonlinear regression models under scale mixtures of skew-normal (SMSN) distributions is developed. We derive a simple EM-type algorithm for iteratively computing maximum likelihood (ML) estimates and the observed information matrix is derived analytically. Simulation studies demonstrate the robustness of this flexible class against outlying and influential observations, as well as nice asymptotic properties of the proposed EM-type ML estimates. Finally, the methodology is illustrated using an ultrasonic calibration data.     

具有马尔可夫骨架的随机过程(英文)

在本文中，我们引入了具有马尔科夫骨架的随机过程的概念．具体讨论了一类重要的具有马氏骨架的随机过程─—（H，Q）─过程，得到了计算这种过程概率分布的向后和向前方程，计算了过程的一维分布以及给出了这种过程正则的充分必要条件．并且本文绘出了文献[3]中主要结果的证明．     

Some concepts of positive dependence for bivariate interchangeable distributions

In this work we consider some familiar and some new concepts of positive dependence for interchangeable bivariate distributions. By characterizing distributions which are positively dependent according to some of these concepts, we indicate real situations in which these concepts arise naturally. For the various families of positively dependent distributions we prove some closure properties and demonstrate all the possible logical relations. Some inequalities are shown and applied to determine whether under- (or over-) estimates, of various probabilistic quantities, occur when a positively dependent distribution is assumed (falsely) to be the product of its marginals (that is, when two positively dependent random variables are assumed, falsely, to be independent). Specific applications in reliability theory, statistical mechanics and reversible Markov processes are discussed. This work was partially supported by National Science Foundation GP-30707X1. It is part of the author's Ph.D. dissertation prepared at the University of Rochester and supervised by A. W. Marshall. Now at Indiana University.     

Asymptotic behavior of posterior distributions for random processes under incorrect models

In this paper, the asymptotic behavior of posterior distributions on parameters contained in random processes is examined when the specified model for the densities is not necessarily correct. Uniform convergence of likelihood functions in some way is shown to be a sufficient condition for the posterior distributions to be asymptotically confined to a set (Theorem 1). For ergodic stationary Markov processes uniform convergence of likelihood functions is established by the ergodic theorem for Banach-valued stationary processes (Proposition 1). A sufficient condition for the uniform convergence is also shown for general random processes (Proposition 2). These results are used to analyze the asymptotic behavior of posterior distributions on parameters contained in linear systems under incorrect models (Example 1 and 2).     

Asymptotic optimal inference for multivariate branching-Markov processes via martingale estimating functions and mixed normality

Multivariate tree-indexed Markov processes are discussed with applications. A Galton-Watson super-critical branching process is used to model the random tree-indexed process. Martingale estimating functions are used as a basic framework to discuss asymptotic properties and optimality of estimators and tests. The limit distributions of the estimators turn out to be mixtures of normals rather than normal. Also, the non-null limit distributions of standard test statistics such as Wald, Rao’s score, and likelihood ratio statistics are shown to have mixtures of non-central chi-square distributions. The models discussed in this paper belong to the local asymptotic mixed normal family. Consequently, non-standard limit results are obtained.     

Ito游程理论与一般Markov链的平稳分布

在本文中,我们利用Ito游程理论给出一般Markov链的平稳分布,该公式包含了极小过程中断和含瞬时态的情形。最后我们给出一个含瞬时态的Markov链的计算例子。     

Existence and stability for Fokker–Planck equations with log-concave reference measure

We study Markov processes associated with stochastic differential equations, whose non-linearities are gradients of convex functionals. We prove a general result of existence of such Markov processes and a priori estimates on the transition probabilities. The main result is the following stability property: if the associated invariant measures converge weakly, then the Markov processes converge in law. The proofs are based on the interpretation of a Fokker–Planck equation as the steepest descent flow of the relative entropy functional in the space of probability measures, endowed with the Wasserstein distance.     

Recursive computation of invariant distributions of Feller processes

This paper provides a general and abstract approach to compute invariant distributions for Feller processes. More precisely, we show that the recursive algorithm presented in Lamberton and Pagès (2002) and based on simulation algorithms of stochastic schemes with decreasing steps can be used to build invariant measures for general Feller processes. We also propose various applications: Approximation of Markov Brownian diffusion stationary regimes with a Milstein or an Euler scheme and approximation of a Markov switching Brownian diffusion stationary regimes using an Euler scheme.     

Optimal training sample allocation and asymptotic expansions for error rates in discriminant analysis

The sample-based rule obtained from Bayes classification rule by replacing the unknown parameters by ML estimates from a stratified training sample is used for the classification of a random observationX into one ofL populations. The asymptotic expansions in terms of the inverses of the training sample sizes for cross-validation, apparent and plug-in error rates are found. These are used to compare estimation methods of the error rate for a wide range of regular distributions as probability models for considered populations. The optimal training sample allocation minimizing the asymptotic expected error regret is found in the cases of widely applicable, positively skewed distributions (Rayleigh and Maxwell distributions). These probability models for populations are often met in ecology and biology. The results indicate that equal training sample sizes for each populations sometimes are not optimal, even when prior probabilities of populations are equal.     

A Hybrid High-Order Markov Chain Model for Computer Intrusion Detection

A hybrid model based mostly on a high-order Markov chain and occasionally on a statistical-independence model is proposed for profiling command sequences of a computer user in order to identify a “signature behavior” for that user. Based on the model, an estimation procedure for such a signature behavior driven by maximum likelihood (ML) considerations is devised. The formal ML estimates are numerically intractable, but the ML-optimization problem can be substituted by a linear inverse problem with positivity constraint (LININPOS), for which the EM algorithm can be used as an equation solver to produce an approximate ML-estimate. The intrusion detection system works by comparing a user's command sequence to the user's and others' estimated signature behaviors in real time through statistical hypothesis testing. A form of likelihood-ratio test is used to detect if a given sequence of commands is from the proclaimed user, with the alternative hypothesis being a masquerader user. Applying the model to real-life data collected from AT&T Labs–Research indicates that the new methodology holds some promise for intrusion detection.     

Ultracontractivity for Markov Semigroups and Quasi-Stationary Distributions

We prove the existence and uniqueness of quasi-stationary distributions for symmetric Markov processes. In particular, we show that if its Markov semigroup is intrinsic ultracontractive, then there exists a unique quasi-stationary distribution. We apply our results to one-dimensional diffusion processes.     

The first-passage times of phase semi-Markov processes

In this paper, we consider a class of semi-Markov processes, known as phase semi-Markov processes, which can be considered as an extension of Markov processes, but whose times between transitions are phase-type random variables. Based on the theory of generalized inverses, we derive expressions for the moments of the first-passage time distributions, generalizing the results obtained by Kemeny and Snell (1960) for Markov chains.     

Optimally Adjusted Mixture Sampling and Locally Weighted Histogram Analysis

Consider the two problems of simulating observations and estimating expectations and normalizing constants for multiple distributions. First, we present a self-adjusted mixture sampling method, which accommodates both adaptive serial tempering and a generalized Wang–Landau algorithm. The set of distributions are combined into a labeled mixture, with the mixture weights depending on the initial estimates of log normalizing constants (or free energies). Then, observations are generated by Markov transitions, and free energy estimates are adjusted online by stochastic approximation. We propose two stochastic approximation schemes by Rao–Blackwellization of the scheme commonly used, and derive the optimal choice of a gain matrix, resulting in the minimum asymptotic variance for free energy estimation, in a simple and feasible form. Second, we develop an offline method, locally weighted histogram analysis, for estimating free energies and expectations, using all the simulated data from multiple distributions by either self-adjusted mixture sampling or other sampling algorithms. This method can be computationally much faster, with little sacrifice of statistical efficiency, than a global method currently used, especially when a large number of distributions are involved. We provide both theoretical results and numerical studies to demonstrate the advantages of the proposed methods.     

Parameter estimation of partially observed continuous time stochastic processes via the EM algorithm

An algorithm is presented for the problem of maximum likelihood (ML) estimation of parameters of partially observed continuous time random processes. This algorithm is an extension of the EM algorithm [3] used in the time series literature, and preserves its main features. It is then applied to the problem of parameter estimation of continuous time, finite state or infinite state (diffusions) Markov processes observed via a noisy sensor. The algorithm in general involves iterations of non-linear smoothing with known parameters and then a non-stochastic maximization. For special cases, including linear models and AR/ARMA processes observed in white noise, each iteration is easily performed with finite dimensional filters. Finally, the algorithm is applied to parameter estimation of “randomly slowly varying” linear systems observed in white noise, and explicit results are derived.