Search results for " optimization."
showing 10 items of 2333 documents
Point process diagnostics based on weighted second-order statistics and their asymptotic properties
2008
A new approach for point process diagnostics is presented. The method is based on extending second-order statistics for point processes by weighting each point by the inverse of the conditional intensity function at the point’s location. The result is generalized versions of the spectral density, R/S statistic, correlation integral and K-function, which can be used to test the fit of a complex point process model with an arbitrary conditional intensity function, rather than a stationary Poisson model. Asymptotic properties of these generalized second-order statistics are derived, using an approach based on martingale theory.
Deriving Reference Decisions
1998
To solve a statistical decision problem from a Bayesian viewpoint, the decision maker must specify a probability distribution on the parameter space, his prior distribution. In order to analyze the influence of this prior distribution on the solution of the problem, Bernardo (1981) proposed to compare the results with those that one would obtain by using that prior distribution which maximizes the useful experimental information, thus introducing the concept of reference decision. This definition is too involved for most of the problems usually found in practice. Here we analyze situations in which it is possible to simplify the definition of the reference decision, and we provide condition…
A more efficient second order blind identification method for separation of uncorrelated stationary time series
2016
The classical second order source separation methods use approximate joint diagonalization of autocovariance matrices with several lags to estimate the unmixing matrix. Based on recent asymptotic results, we propose a novel unmixing matrix estimator which selects the best lag set from a finite set of candidate sets specified by the user. The theory is illustrated by a simulation study.
Bayesian Smoothing in the Estimation of the Pair Potential Function of Gibbs Point Processes
1999
A flexible Bayesian method is suggested for the pair potential estimation with a high-dimensional parameter space. The method is based on a Bayesian smoothing technique, commonly applied in statistical image analysis. For the calculation of the posterior mode estimator a new Monte Carlo algorithm is developed. The method is illustrated through examples with both real and simulated data, and its extension into truly nonparametric pair potential estimation is discussed.
Model selection in linear mixed-effect models
2019
Linear mixed-effects models are a class of models widely used for analyzing different types of data: longitudinal, clustered and panel data. Many fields, in which a statistical methodology is required, involve the employment of linear mixed models, such as biology, chemistry, medicine, finance and so forth. One of the most important processes, in a statistical analysis, is given by model selection. Hence, since there are a large number of linear mixed model selection procedures available in the literature, a pressing issue is how to identify the best approach to adopt in a specific case. We outline mainly all approaches focusing on the part of the model subject to selection (fixed and/or ra…
Parallel Construction and Query of Index Data Structures for Pattern Matching on Square Matrices
1999
AbstractWe describe fast parallel algorithms for building index data structures that can be used to gather various statistics on square matrices. The main data structure is the Lsuffix tree, which is a generalization of the classical suffix tree for strings. Given ann×ntext matrixA, we build our data structures inO(logn) time withn2processors on a CRCW PRAM, so that we can quickly processAin parallel as follows: (i) report some statistical information aboutA, e.g., find the largest repeated square submatrices that appear at least twice inAor determine, for each position inA, the smallest submatrix that occurs only there; (ii) given, on-line, anm×mpattern matrixPAT, check whether it occurs i…
ConvergenceClubs: A Package for Performing the Phillips and Sul's Club Convergence Clustering Procedure
2019
This paper introduces package ConvergenceClubs, which implements functions to perform the Phillips and Sul (2007, 2009) club convergence clustering procedure in a simple and reproducible manner. The approach proposed by Phillips and Sul to analyse the convergence patterns of groups of economies is formulated as a nonlinear time varying factor model that allows for different time paths as well as individual heterogeneity. Unlike other approaches in which economies are grouped a priori, it also allows the endogenous determination of convergence clubs. The algorithm, usage, and implementation details are discussed.
Optimal designs for a one-way layout with covariates
2000
Abstract For the general class of Φ q -criteria optimal designs are characterized which reflect the inherent symmetry in a one-way layout with covariates. In particular, the eigenvalues of the covariance matrices are related to those in suitably chosen marginal models depending on the underlying interaction structure.
Linear and ellipsoidal restrictions in linear regression
1991
The problem of combining linear and ellipsoidal restrictions in linear regression is investigated. Necessary and sufficient conditions for compactness of the restriction set are proved assuring the existence of a minimax estimator. When the restriction set is not compact a minimax estimator may still exist for special loss functions arid regression designs
Bayesian hierarchical models in manufacturing bulk service queues
2006
In this paper, Queueing Theory and Bayesian statistical tools are used to analyze the congestion of various manufacturing bulk service queues with the same characteristics that are working independently of one another and in equilibrium. Hierarchical models are discussed in order to develop the whole inferential process for the parameters governing the system. Markov Chain Monte Carlo methods and numerical inversion of transforms are addressed to compute the posterior predictive distributions of the usual measures of performance in practice.