Abstract not found

Long term yield estimates for natural forests require a harvesting model to enable future yields to be estimated reliably. The model should predict the felled stems, the proportion of these which are merchantable, and any damage to the residual stand. Regression analyses was used to develop a model of current logging practice in the rain forests of north Queensland. Logistic functions predict the probability of any tree being marked for logging, the probability of a felled tree being merchantable, and the probability of any tree in the residual stand being damaged by logging. Important predictor variables included tree species and size, merchantable basal area, basal area logged, logging history, and topography. There was no evidence to suggest that soil type or site quality influenced current treemarking practice. The approach is applicable to other mixed forest types managed for selection logging.

We discuss the problem of producing an epitome, or brief summary, of a Bayesian posterior distribution - and then investigate a general solution based on the Minimum Message Length (MML) principle. Clearly, the optimal criterion for choosing such an epitome is determined by the epitome's intended use. The interesting general case is where this use is unknown since, in order to be practical, the choice of epitome criterion becomes subjective. We identify a number of desirable properties that an epitome could have - facilitation of point estimation, human comprehension, and fast approximation of posterior expectations. We call these the properties of Bayesian Posterior Comprehension and show that the Minimum Message Length principle can be viewed as an epitome criterion that produces epitomes having these properties. We then present and extend Message from Monte Carlo as a means for constructing instantaneous Minimum Message Length codebooks (and epitomes) using Markov Chain Monte Carlo methods. The Message from Monte Carlo methodology is illustrated for binary regression, generalised linear model, and multiple change-point problems.

Abstract—Applications are given of a formula for the exact probability density function of the maximum likelihood estimates of a statistical model, where the data generating model is allowed to differ from the estimation model. The main examples are supported by simulation experiments. Curved exponential families are investigated, for which an approach is described that can be used in many practical situations. The distribution of a maximum likelihood estimator in exponential regression is developed. Nonlinear regression is then considered, with an example of a model discrepancy situation arising in ELISA immunoassays and similar biochemical titrations. An incorrect logistic model is specified for a titration curve that is used for describing the reaction of a chemical sample to applied substrate concentration. A method is suggested to reduce the amount of bias in the estimate of binding affinity. Finally there is a prospective discussion of other possible uses of the technique, including general comparisons of sets of alternative models in frequentist and Bayesian settings, applications to robust estimation and extensions beyond maximum likelihood estimates.

The indicator value and ecological amplitude of a species with respect to a quantitative environmental vari-able can be estimated from data on species occurrence and environment. A simple weighted averaging (WA) method for estimating these parameters is compared by simulation with the more elaborate method of Gaus-sian logistic regression (GLR), a form of the generalized linear model which fits a Gaussian-like species re-sponse curve to presence-absence data. The indicator value and the ecological amplitude are expressed by two parameters of this curve, termed the optimum and the tolerance, respectively. When a species is rare and has a narrow ecological amplitude- or when the distribution of quadrats along the environmental variable is reasonably even over the species ' range, and the number of quadrats is small- then WA is shown to ap-proach GLR in efficiency. Otherwise WA may give misleading results. GLR is therefore preferred as a practi-cal method for summarizing species ' distributions along environmental gradients. Formulas are given to cal-culate species optima and tolerances (with their standard errors), and a confidence interval for the optimum from the GLR output of standard statistical packages.

Introduction and Disclaimer These worked examples illustrate the use of the BUGS language and sampler in a wide range of problems. They contain a number of useful "tricks", but are certainly not exhaustive of the models that may be analysed. We emphasise that all the results for these examples have been derived in the most naive way: in general a burn-in of 500 iterations and a single long run of 1000 iterations. This is not recommended as a general technique: no tests of convergence have been carried out, and traces of the estimates have not even been plotted. However, comparisons with published results have been made where possible. Times have been measured on a 60 MHz superSPARC: a 60 MHz Pentium PC appears to be about 4 times slower, and a 30 MHz superSPARC about 2 times slower. Users are warned to be extremely careful about assuming convergence, especially when using complex models including errors in variables, crossed random effects and intrinsic p

Introduction and Disclaimer These worked examples illustrate the use of the BUGS language and sampler in a wide range of problems. They contain a number of useful "tricks", but are certainly not exhaustive of the models that may be analysed. We emphasise that all the results for these examples have been derived in the most naive way: in general a burn-in of 500 iterations and a single long run of 1000 iterations. This is not recommended as a general technique: no tests of convergence have been carried out, and traces of the estimates have not even been plotted. However, comparisons with published results have been made where possible. Times have been measured on a 60 MHz superSPARC: a 60 MHz Pentium PC appears to be about 4 times slower, and a 30 MHz superSPARC about 2 times slower. Users are warned to be extremely careful about assuming convergence, especially when using complex models including errors in variables, crossed random effects and intrinsi

Abstract—Applications are given of a formula for the exact probability density function of the maximum likelihood estimates of a statistical model, where the data generating model is allowed to differ from the estimation model. The main examples are supported by simulation experiments. Curved exponential families are investigated, for which an approach is described that can be used in many practical situations. The distribution of a maximum likelihood estimator in exponential regression is developed. Nonlinear regression is then considered, with an example of a model discrepancy situation arising in ELISA immunoassays and similar biochemical titrations. An incorrect logistic model is specified for a titration curve that is used for describing the reaction of a chemical sample to applied substrate concentration. A method is suggested to reduce the amount of bias in the estimate of binding affinity. There is a discussion of other possible uses for the technique.

Recognizing and foreseeing which credit clients will be "good orbad payers " is an important and di cult task for bank institutions and credit protection services. Using data from approximately 10,000 clients obtained from a large private Brazilian bank, we present a methodology to perform the credit scoring analysis. The methodology proposed is divided into 2 stages: statistical data analysis and the use of a model to perform the Pattern Recognition, discriminating the two groups mentioned earlier. Keywords: Pattern Recognition, Credit Scoring, Multivariate Analysis.

generalised linear models