Issues of robustness and high dimensionality in cluster analysis

Finite mixture models are being increasingly used to model the distributions of a wide variety of random phenomena. While normal mixture models are often used to cluster data sets of continuous multivariate data, a more robust clustering can be obtained by considering the t mixture model-based approach. Mixtures of factor analyzers enable model-based density estimation to be undertaken for high-dimensional data where the number of observations n is very large relative to their dimension p. As the approach using the multivariate normal family of distributions is sensitive to outliers, it is more robust to adopt the multivariate t family for the component error and factor distributions. The computational aspects associated with robustness and high dimensionality in these approaches to cluster analysis are discussed and illustrated.

A rough cluster analysis of shopping orientation data

This paper describes the application of a new technique, rough clustering, to the problem of market segmentation. Rough clustering produces different solutions to k-means analysis because of the possibility of multiple cluster membership of objects. Traditional clustering methods generate extensional descriptions of groups, that show which objects are members of each cluster. Clustering techniques based on rough sets theory generate intensional descriptions, which outline the main characteristics of each cluster. In this study, a rough cluster analysis was conducted on a sample of 437 responses from a larger study of the relationship between shopping orientation (the general predisposition of consumers toward the act of shopping) and intention to purchase products via the Internet. The cluster analysis was based on five measures of shopping orientation: enjoyment, personalization, convenience, loyalty, and price. The rough clusters obtained provide interpretations of different shopping orientations present in the data without the restriction of attempting to fit each object into only one segment. Such descriptions can be an aid to marketers attempting to identify potential segments of consumers.

O NEGÓCIO DA COMUNICAÇÃO EM TURISMO Cluster no Núcleo de Turismo da Editora Abril

Esse trabalho examina a relação da Comunicação com o Turismo discutindo a importância dos produtos de comunicação junto ao negócio do turismo no Brasil. Tem como corpus da pesquisa o Núcleo de Turismo da Editora Abril, formado pelas publicações Guia 4 Rodas, revistas Viagem e Turismo e National Geographic Brasil, e o portal de viagens www.viajeaqui.com.br. Investiga o peso e o amadurecimento do setor de turismo na economia brasileira, o crescimento e a consolidação do mercado editorial neste segmento. Discute o papel do Cluster na área de Turismo e seus efeitos no Jornalismo. Por meio da análise de conteúdo avalia os processos de produção das publicações do Núcleo, o caminho das informações compartilhadas pelos profissionais envolvidos e o impacto dessas informações nos públicos-alvos de cada veículo. Faz ainda uma reflexão sobre os limites entre Jornalismo e Publicidade em mídias segmentadas dentro de um Cluster.(AU)

The identification of pathological subtypes of Alzheimer's disease using cluster analysis

A culster analysis was performed on 78 cases of Alzheimer's disease (AD) to identify possible pathological subtypes of the disease. Data on 47 neuropathological variables, inculding features of the gross brain and the density and distribution of senile plaques (SP) and neurofibrillary tangles (NFT) were used to describe each case. Cluster analysis is a multivariate statistical method which combines together in groups, AD cases with the most similar neuropathological characteristics. The majority of cases (83%) were clustered into five such groups. The analysis suggested that an initial division of the 78 cases could be made into two major groups: (1) a large group (68%) in which the distribution of SP and NFT was restricted to a relatively small number of brain regions, and (2) a smaller group (15%) in which the lesions were more widely disseminated throughout the neocortex. Each of these groups could be subdivided on the degree of capillary amyloid angiopathy (CAA) present. In addition, those cases with a restricted development of SP/NFT and CAA could be divided further into an early and a late onset form. Familial AD cases did not cluster as a separate group but were either distributed between four of the five groups or were cases with unique combinations of pathological features not closely related to any of the groups. It was concluded that multivariate statistical methods may be of value in the classification of AD into subtypes. © 1994 Springer-Verlag.

Classic beta-amyloid deposits cluster around large diameter blood vessels rather than capillaries in sporadic Alzheimer's disease

Various hypotheses could explain the relationship between beta-amyloid (Abeta) deposition and the vasculature in Alzheimer's disease (AD). Amyloid deposition may reduce capillary density, affect endothelial cells of blood vessels, result in diffusion from blood vessels, or interfere with the perivascular clearance mechanism. Hence, the spatial pattern of the classic ('cored') type of Abeta deposit was studied in the upper laminae (I,II/III) of the superior frontal gyrus in nine cases of sporadic AD (SAD). Sections were immunostained with antibodies against Abeta and with collagen IV to study the relationships between the spatial distribution of the classic deposits and the blood vessel profiles. Both the classic deposits and blood vessel profiles were distributed in clusters. In all cases, there was a positive spatial correlation between the clusters of the classic deposits and the larger diameter (>10 microm) blood vessel profiles and especially the vertically penetrating arterioles. In only 1 case, was there a significant spatial correlation between the clusters of the classic deposits and the smaller diameter (<10 microm) capillaries. There were no negative correlations between the density of Abeta deposits and the smaller diameter capillaries. In 9/11 cases, the clusters of the classic deposits were significantly larger than those of the clusters of the larger blood vessel profiles. In addition, the density of the classic deposits declined as a negative exponential function with distance from a vertically penetrating arteriole. These results suggest that the classic Abeta deposits cluster around the larger blood vessels in the upper laminae of the frontal cortex. This aggregation could result from diffusion of proteins from blood vessels or from overloading the system of perivascular clearance from the brain.

The use of multivariate methods in the identification of subtypes of Alzheimer's disease: a comparison of principal components and cluster analysis

Two contrasting multivariate statistical methods, viz., principal components analysis (PCA) and cluster analysis were applied to the study of neuropathological variations between cases of Alzheimer's disease (AD). To compare the two methods, 78 cases of AD were analyzed, each characterised by measurements of 47 neuropathological variables. Both methods of analysis revealed significant variations between AD cases. These variations were related primarily to differences in the distribution and abundance of senile plaques (SP) and neurofibrillary tangles (NFT) in the brain. Cluster analysis classified the majority of AD cases into five groups which could represent subtypes of AD. However, PCA suggested that variation between cases was more continuous with no distinct subtypes. Hence, PCA may be a more appropriate method than cluster analysis in the study of neuropathological variations between AD cases.

A study of cluster analysis techniques and their applications

This thesis seeks to describe the development of an inexpensive and efficient clustering technique for multivariate data analysis. The technique starts from a multivariate data matrix and ends with graphical representation of the data and pattern recognition discriminant function. The technique also results in distances frequency distribution that might be useful in detecting clustering in the data or for the estimation of parameters useful in the discrimination between the different populations in the data. The technique can also be used in feature selection. The technique is essentially for the discovery of data structure by revealing the component parts of the data. lhe thesis offers three distinct contributions for cluster analysis and pattern recognition techniques. The first contribution is the introduction of transformation function in the technique of nonlinear mapping. The second contribution is the us~ of distances frequency distribution instead of distances time-sequence in nonlinear mapping, The third contribution is the formulation of a new generalised and normalised error function together with its optimal step size formula for gradient method minimisation. The thesis consists of five chapters. The first chapter is the introduction. The second chapter describes multidimensional scaling as an origin of nonlinear mapping technique. The third chapter describes the first developing step in the technique of nonlinear mapping that is the introduction of "transformation function". The fourth chapter describes the second developing step of the nonlinear mapping technique. This is the use of distances frequency distribution instead of distances time-sequence. The chapter also includes the new generalised and normalised error function formulation. Finally, the fifth chapter, the conclusion, evaluates all developments and proposes a new program. for cluster analysis and pattern recognition by integrating all the new features.