Structured Spatio-Temporal Shot-Noise Cox Point Process Models, with a View to Modelling Forest Fires

Abstract.  Spatio-temporal Cox point process models with a multiplicative structure for the driving random intensity, incorporating covariate information into temporal and spatial components, and with a residual term modelled by a shot-noise process, are considered. Such models are flexible and tractable for statistical analysis, using spatio-temporal versions of intensity and inhomogeneous K -functions, quick estimation procedures based on composite likelihoods and minimum contrast estimation, and easy simulation techniques. These advantages are demonstrated in connection with the analysis of a relatively large data set consisting of 2796 days and 5834 spatial locations of fires. The model is compared with a spatio-temporal log-Gaussian Cox point process model, and likelihood-based methods are discussed to some extent.     

ASYMPTOTIC TESTS FOR THE CONTINUOUS TIME MARKOV BRANCHING PROCESS WITH IMMIGRATION

This paper presents limit distributions for the modified score and the likelihood-ratio (LR) statistic for testing a composite hypothesis involving the split intensity and mean of the offspring distribution of the supercritical continuous time Markov branching process allowing immigration (CBPI). The immigration intensity and mean are treated as nuisance parameters.     

Power Loss for Inhomogeneous Poisson Processes

In this work, based on a realization of an inhomogeneous Poisson process whose intensity function depends on a real parameter, we consider a simple null hypothesis against the composite one sided alternative. Under certain regularity conditions we will obtain the power loss of the score test which measures its performance with respect to the Neyman-Pearson test. We present the second-order approximation of the power of the score test under the close alternatives by specifying the explicit form of the next term after the Gaussian term.     

Two step estimation for Neyman-Scott point process with inhomogeneous cluster centers

This paper is concerned with parameter estimation for the Neyman-Scott point process with inhomogeneous cluster centers. Inhomogeneity depends on spatial covariates. The regression parameters are estimated at the first step using a Poisson likelihood score function. Three estimation procedures (minimum contrast method based on a modified K function, composite likelihood and Bayesian methods) are introduced for estimation of clustering parameters at the second step. The performance of the estimation methods are studied and compared via a simulation study. This work has been motivated and illustrated by ecological studies of fish spatial distribution in an inland reservoir.     

Residual analysis for spatial point processes (with discussion)

Summary.  We define residuals for point process models fitted to spatial point pattern data, and we propose diagnostic plots based on them. The residuals apply to any point process model that has a conditional intensity; the model may exhibit spatial heterogeneity, interpoint interaction and dependence on spatial covariates. Some existing ad hoc methods for model checking (quadrat counts, scan statistic, kernel smoothed intensity and Berman's diagnostic) are recovered as special cases. Diagnostic tools are developed systematically, by using an analogy between our spatial residuals and the usual residuals for (non-spatial) generalized linear models. The conditional intensity λ plays the role of the mean response. This makes it possible to adapt existing knowledge about model validation for generalized linear models to the spatial point process context, giving recommendations for diagnostic plots. A plot of smoothed residuals against spatial location, or against a spatial covariate, is effective in diagnosing spatial trend or co-variate effects. Q – Q -plots of the residuals are effective in diagnosing interpoint interaction.     

Tests for the Proportional Intensity Assumption Based on the Score Process

Proportional intensity models are widely used for describing the relationship between the intensity of a counting process and associated covariates. A basic assumption in this model is the proportionality, that each covariate has a multiplicative effect on the intensity. We present and study tests for this assumption based on a score process which is equivalent to cumulative sums of the Schoenfeld residuals. Tests within principle power against any type of departure from proportionality can be constructed based on this score process. Among the tests studied, in particular an Anderson-Darling type test turns out to be very useful by having good power properties against general alternatives. A simulation study comparing various tests for proportionality indicates that this test seems to be a good choice for an omnibus test for proportionality.     

Rao's score test in spatial econometrics

Rao's score test provides an extremely useful framework for developing diagnostics against hypotheses that reflect cross-sectional or spatial correlation in regression models, a major focus of attention in spatial econometrics. In this paper, a review and assessment is presented of the application of Rao's score test against three broad classes of spatial alternatives: spatial autoregressive and moving average processes, spatial error components and direct representation models. A brief review is presented of the various forms and distinctive characteristics of RS tests against spatial processes. New tests are developed against the alternatives of spatial error components and direct representation models. It is shown that these alternatives do not conform to standard regularity conditions for maximum likelihood estimation. In the case of spatial error components, the RS test does have the standard asymptotic properties, whereas Wald and Likelihood Ratio tests do not. Direct representation models yield a situation where the nuisance parameter is only identified under the alternative, such that a Davies-type approximation to the significance level of the RS test is necessary. The performance of both new RS tests is illustrated in a small number of Monte Carlo simulation experiments.     

Multivariate testing and model-checking for generalized order statistics with applications

The exponential family structure of the joint distribution of generalized order statistics is utilized to establish multivariate tests on the model parameters. For simple and composite null hypotheses, the likelihood ratio test (LR test), Wald's test, and Rao's score test are derived and turn out to have simple representations. The asymptotic distribution of the corresponding test statistics under the null hypothesis is stated, and, in case of a simple null hypothesis, asymptotic optimality of the LR test is addressed. Applications of the tests are presented; in particular, we discuss their use in reliability, and to decide whether a Poisson process is homogeneous. Finally, a power study is performed to measure and compare the quality of the tests for both, simple and composite null hypotheses.     

Information criteria for inhomogeneous spatial point processes

The theoretical foundation for a number of model selection criteria is established in the context of inhomogeneous point processes and under various asymptotic settings: infill, increasing domain and combinations of these. For inhomogeneous Poisson processes we consider Akaike's information criterion and the Bayesian information criterion, and in particular we identify the point process analogue of ‘sample size’ needed for the Bayesian information criterion. Considering general inhomogeneous point processes we derive new composite likelihood and composite Bayesian information criteria for selecting a regression model for the intensity function. The proposed model selection criteria are evaluated using simulations of Poisson processes and cluster point processes.     

Approximate Bayesian computation with composite score functions

Both approximate Bayesian computation (ABC) and composite likelihood methods are useful for Bayesian and frequentist inference, respectively, when the likelihood function is intractable. We propose to use composite likelihood score functions as summary statistics in ABC in order to obtain accurate approximations to the posterior distribution. This is motivated by the use of the score function of the full likelihood, and extended to general unbiased estimating functions in complex models. Moreover, we show that if the composite score is suitably standardised, the resulting ABC procedure is invariant to reparameterisations and automatically adjusts the curvature of the composite likelihood, and of the corresponding posterior distribution. The method is illustrated through examples with simulated data, and an application to modelling of spatial extreme rainfall data is discussed.     

On weighted composite scores for early Alzheimer's trials

Recent research on finding appropriate composite endpoints for preclinical Alzheimer's disease has focused considerable effort on finding “optimized” weights in the construction of a weighted composite score. In this paper, several proposed methods are reviewed. Our results indicate no evidence that these methods will increase the power of the test statistics, and some of these weights will introduce biases to the study. Our recommendation is to focus on identifying more sensitive items from clinical practice and appropriate statistical analyses of a large Alzheimer's data set. Once a set of items has been selected, there is no evidence that adding weights will generate more sensitive composite endpoints.     

Hypotheses testing for generalized order statistics with simple order restrictions on model parameters under the alternative

In applications of generalized order statistics as, for instance, reliability analysis of engineering systems, prior knowledge about the order of the underlying model parameters is often available and may therefore be incorporated in inferential procedures. Taking this information into account, we establish the likelihood ratio test, Rao's score test, and Wald's test for test problems arising from the question of appropriate model selection for ordered data, where simple order restrictions are put on the parameters under the alternative hypothesis. For simple and composite null hypothesis, explicit representations of the corresponding test statistics are obtained along with some properties and their asymptotic distributions. A simulation study is carried out to compare the order restricted tests in terms of their power. In the set-up considered, the adapted tests significantly improve the power of the associated omnibus versions for small sample sizes, especially when testing a composite null hypothesis.     

A functional data approach to model score difference process in professional basketball games

In this paper, we investigate the progress of score difference (between home and away teams) in professional basketball games employing functional data analysis (FDA). The observed score difference is viewed as the realization of the latent intensity process, which is assumed to be continuous. There are two major advantages of modeling the latent score difference intensity process using FDA: (1) it allows for arbitrary dependent structure among score change increments. This removes potential model mis-specifications and accommodates momentum which is often observed in sports games. (2) further statistical inferences using FDA estimates will not suffer from inconsistency due to the issue of having a continuous model yet discretely sampled data. Based on the FDA estimates, we define and numerically characterize momentum in basketball games and demonstrate its importance in predicting game outcomes.     

Practical Maximum Pseudolikelihood for Spatial Point Patterns(with Discussion)

This paper describes a technique for computing approximate maximum pseudolikelihood estimates of the parameters of a spatial point process. The method is an extension of Berman & Turner's (1992) device for maximizing the likelihoods of inhomogeneous spatial Poisson processes. For a very wide class of spatial point process models the likelihood is intractable, while the pseudolikelihood is known explicitly, except for the computation of an integral over the sampling region. Approximation of this integral by a finite sum in a special way yields an approximate pseudolikelihood which is formally equivalent to the (weighted) likelihood of a loglinear model with Poisson responses. This can be maximized using standard statistical software for generalized linear or additive models, provided the conditional intensity of the process takes an 'exponential family' form. Using this approach a wide variety of spatial point process models of Gibbs type can be fitted rapidly, incorporating spatial trends, interaction between points, dependence on spatial covariates, and mark information.     

Bayesian mixture modeling for spatial Poisson process intensities,with applications to extreme value analysis

We propose a method for the analysis of a spatial point pattern, which is assumed to arise as a set of observations from a spatial nonhomogeneous Poisson process. The spatial point pattern is observed in a bounded region, which, for most applications, is taken to be a rectangle in the space where the process is defined. The method is based on modeling a density function, defined on this bounded region, that is directly related with the intensity function of the Poisson process. We develop a flexible nonparametric mixture model for this density using a bivariate Beta distribution for the mixture kernel and a Dirichlet process prior for the mixing distribution. Using posterior simulation methods, we obtain full inference for the intensity function and any other functional of the process that might be of interest. We discuss applications to problems where inference for clustering in the spatial point pattern is of interest. Moreover, we consider applications of the methodology to extreme value analysis problems. We illustrate the modeling approach with three previously published data sets. Two of the data sets are from forestry and consist of locations of trees. The third data set consists of extremes from the Dow Jones index over a period of 1303 days.     

Globally intensity-reweighted estimators for K- and pair correlation functions

We introduce new estimators of the inhomogeneous K-function and the pair correlation function of a spatial point process as well as the cross K-function and the cross pair correlation function of a bivariate spatial point process under the assumption of second-order intensity-reweighted stationarity. These estimators rely on a ‘global’ normalisation factor which depends on an aggregation of the intensity function, while the existing estimators depend ‘locally’ on the intensity function at the individual observed points. The advantages of our new global estimators over the existing local estimators are demonstrated by theoretical considerations and a simulation study.     

Large Sample Inference for Markovian Exponential Families with Application to Branching Processes with Immigration

This paper presents limit distributions for the score and likelihood-ratio (L.R.) statistic for testing a composite hypothesis involving the mean of the offspring distribution of the Bienaymé-Galton-Watson branching process with immigration (BPWI) when the process is subcritical, critical or supercritical. The BPWI is shown to be a member of a certain Markovian exponential family.     

Spatio-temporal improvised explosive device monitoring: improving detection to minimise attacks

The improvised explosive device (IED) is a weapon of strategic influence on today's battlefield. IED detonations occur predominantly on roads, footpaths, or trails. Therefore, locations are best described when constrained to the road network, and some spaces on the network are more dangerous at specific times of the day. We propose a statistical model that reduces the spatial location to one dimension and uses a cyclic time as a second dimension. Based on the Poisson process methodology, we develop normalised, inhomogeneous, bivariate intensity functions measuring the threat of attack to support resourcing decisions. A simulation and an analysis of attacks on a main supply route in Baghdad are given to illustrate the proposed methods. Additionally, we provide an overview of the growing demand for the analysis efforts in support of operations in Afghanistan and Iraq, and provide an extensive literature review of developments in counter-IED analysis.     

Two-step estimation for inhomogeneous spatial point processes

Summary.  The paper is concerned with parameter estimation for inhomogeneous spatial point processes with a regression model for the intensity function and tractable second-order properties ( K -function). Regression parameters are estimated by using a Poisson likelihood score estimating function and in the second step minimum contrast estimation is applied for the residual clustering parameters. Asymptotic normality of parameter estimates is established under certain mixing conditions and we exemplify how the results may be applied in ecological studies of rainforests.     

Consistent parametric estimation of the intensity of a spatial–temporal point process

We consider conditions under which parametric estimates of the intensity of a spatial–temporal point process are consistent. Although the actual point process being estimated may not be Poisson, an estimate involving maximizing a function that corresponds exactly to the log-likelihood if the process is Poisson is consistent under certain simple conditions. A second estimate based on weighted least squares is also shown to be consistent under quite similar assumptions. The conditions for consistency are simple and easily verified, and examples are provided to illustrate the extent to which consistent estimation may be achieved. An important special case is when the point processes being estimated are in fact Poisson, though other important examples are explored as well.