A Bayesian method for classification and discrimination

We discuss Bayesian analyses of traditional normal-mixture models for classification and discrimination. The development involves application of an iterative resampling approach to Monte Carlo inference, commonly called Gibbs sampling, and demonstrates routine application. We stress the benefits of exact analyses over traditional classification and discrimination techniques, including the ease with which such analyses may be performed in a quite general setting, with possibly several normal-mixture components having different covariance matrices, the computation of exact posterior classification probabilities for observed data and for future cases to be classified, and posterior distributions for these probabilities that allow for assessment of second-level uncertainties in classification.     

Improving performance? The adoption and implementation of quality systems in U.K. nonprofits

There are increasing and multiple pressures on nonprofit organizations to demonstrate excellence in performance. Although there is a growing literature on the various approaches to performance improvement taken by nonprofits, little is known about the processes involved in the adoption and implementation of specific approaches. This article is about the adoption and use of one approach to performance improvement, “quality systems,” in the U.K. nonprofit sector. We report findings about factors that encourage nonprofits to adopt quality systems. We also analyze the distinctive challenges of implementing quality approaches in a nonprofit sector context and suggest critical success factors. The article concludes with a discussion of the organizational and policy implications of applying the management concepts of quality and performance to the nonprofit sector.     

Domains of convergence for the EM algorithm: a cautionary tale in a location estimation problem

The EM algorithm is a popular method for maximizing a likelihood in the presence of incomplete data. When the likelihood has multiple local maxima, the parameter space can be partitioned into domains of convergence, one for each local maximum. In this paper we investigate these domains for the location family generated by the t-distribution. We show that, perhaps somewhat surprisingly, these domains need not be connected sets. As an extreme case we give an example of a domain which consists of an infinite union of disjoint open intervals. Thus the convergence behaviour of the EM algorithm can be quite sensitive to the starting point.     

Class analysis and its futures

Books reviewed in this articles:
Class and Stratification: An introduction to current debates R. Crompton
Class S. Edgell
Changing Classes: Stratification and Mobility in Post-Industrial Societies G. Esping-Anderson, (ed.)
Class and Inequality: Comparative Perspectives M. Hamilton and M. Hirszowicz
Class R. Scase     

Is Stakeholder Corporate Governance Appropriate in Russia?

Journal of Management and Governance - Alternative forms of disciplining the actionsof senior enterprise managers can becharacterized as shareholder or stakeholdergovernance. This paper analyzes...