Knowledge representation by schemata in financial expert systems

This paper introduces the notion of knowledge structuring by schemata. A schema is a chunk of knowledge which generalizes the idea of functional dependence between pertinent concepts (objects). The rules are closely related to the schemata. Moreover the schemata are organized in a hierarchical network which takes into account the relative importance of the concepts. Examples are given in the financial field and the consequences with regards to knowledge engineering are discussed.     

The Social Impact of Industrial and Resource Development on Local Communities: A Survey of Residents' Views

This paper describes the results of two surveys of public attitudes, perceptions, and expectations regarding large-scale energy development projects in the Latrobe Valley, Victoria. A baseline survey was conducted in 1974, before plans for a massive increase in planned energy generation for the Valley had begun to be instituted (projected to inject 440MW per year additional capacity to the year 1989, compared with an average annual increase of 52MW per year which characterized the period from 1921 to 1974 (State Electricity Commission of Victoria 1981)) and before construction commenced in 1977, on the first of two 2000MW power stations, located at Loy Yang, six kilometres from the town of Traralgon in the Latrobe Valley, Victoria. A subsequent survey was undertaken in 1981, with construction well-advanced on the first two 500MW units. The community studies represent part of a major project concerned with impact assessment of new industrial and resource projects in settled (as distinct from isolated) regions of Australia.     

The regulation of family involvement in deinstitutionalization

Conflicting legal, administrative, and professional policies surround family involvement in the deinstitutionalization process. This article addresses these discrepancies and makes recommendations for developing effective and consistent channels of policy among social workers and other helping professionals.     

Estimators of percentiles based on absolute loss

Estimators of percentiles of location and scale parameter distributions are optimized based on Pitman closeness and absolute risk. A median unbiased (MU) estimator and a minimum risk (MR) estimator are shown to exist within a class of estimators, and to depend upon the medians of two completely specified distributions.     

Modelling change in cognitive understanding with finite mixtures

Psychological theories often posit the existence of several different states. Individuals are viewed as belonging to one of the states at a given age, but with development pass to another state. A main problem in evaluating such theories is representing the transition from one state to another over age. A stochastic transition framework is proposed which should be useful in many different settings. The model is illustrated with data from a cognitive development task.     

Saddlepoint-Based Bootstrap Inference for Quadratic Estimating Equations

Abstract.  We propose an easy to implement method for making small sample parametric inference about the root of an estimating equation expressible as a quadratic form in normal random variables. It is based on saddlepoint approximations to the distribution of the estimating equation whose unique root is a parameter's maximum likelihood estimator (MLE), while substituting conditional MLEs for the remaining (nuisance) parameters. Monotoncity of the estimating equation in its parameter argument enables us to relate these approximations to those for the estimator of interest. The proposed method is equivalent to a parametric bootstrap percentile approach where Monte Carlo simulation is replaced by saddlepoint approximation. It finds applications in many areas of statistics including, nonlinear regression, time series analysis, inference on ratios of regression parameters in linear models and calibration. We demonstrate the method in the context of some classical examples from nonlinear regression models and ratios of regression parameter problems. Simulation results for these show that the proposed method, apart from being generally easier to implement, yields confidence intervals with lengths and coverage probabilities that compare favourably with those obtained from several competing methods proposed in the literature over the past half-century.