On the asymptotic properties of fuzzy c-means cluster prototypes as estimators of mixture subpopulation centers

Several computational studies suggest that the fuzzy c-means (FCM) clustering scheme may be used successfully in some cases to obtain estimates for the parameters of a statistical mixture (e.g., for a mixture of normal distributions). While these (limited) simulation results for the fuzzy c-means approach support this hypothesis, we provide herein an example that shows that the FCM cluster prototypes cannot generally be statistically consistent estimators of the centers (means) of any univariate mixture having symmetric component distributions     

A note on the asymptotic distribution of mardia's measure of multivariate kurtosis

Weak convergence results are used to investigate asymptotic properties of Mardia's measure of multivariate kurtosis in the context of assessing multivariate normality.     

Tests concerning two non-isotropic principal components

Kshirsagar & Gupta (1965) proposed test criteria for testing the null hypothesis that a p x p covariance matrix, with known smallest latent root of multiplicity p-2. has its two non-isotropic principal components in specified directions. Simula-tion results indicate that the power functions of these test criteria lack some desirable properties. However, another statis-tic discussed by Kshirsagar and Gupta is shown to have a power-function which performs quite well. The situation in which the smallest latent root of multiplicity p-2 is unknown, is also considered.     

Model specification tests against non-nested alternatives

Non-nested hypothesis tests provide a way to test the specification of an econometric model against the evidence provided by one or more non-nested alternatives. This paper surveys the recent literature on non-nested hypothesis testing in the context of regression and related models. Much of the purely statistical 1iterature which has evolved from the fundamental work of Cox (1961, 1962) is discussed briefly or not at all. Instead, emphasis is placed on those techniques which are easy to employ in practice and are likely to be useful to applied workers.     

An Introduction to Modern Matrix Methods and Statistics

In this article, using the representation that the Kalman filter recursions in state-space models can be expressed as a matrix-weighted average of prior and sample estimates, we supplement the usual filtering algorithm by an extreme bounds analysis. Specifically, as the covariance matrix of the state error is varied in the class of symmetric and positive-definite matrices, the filtering estimates are shown to be in an ellipsoid.     

Zero Correlation,Independence, and Normality

Two easy classroom examples are given of marginally normal random variables that are dependent but have zero correlation. One example is similar in nature to Anderson's well-known illustration that marginal normality does not imply joint normality.     

The role of leadership in shared mental model convergence and team performance improvement: An agent-based computational model

Research in shared mental models has immeasurably aided our understanding of effective teamwork and taskwork. However, little research has focused on the role that leaders play, if any, in influencing, developing and/or fostering shared mental models and thereby improving team performance. We developed an agent-based computational model based on McComb's theory of three-phase mental model development, where agents repeatedly share individual opinions (orientation phase), evaluate and respond to the opinions expressed by others (differentiation phase), and modify their understanding of the team based on the responses (integration phase). Leadership and team properties are represented in three experimental parameters: social network structure, heterogeneity of agents' domains of expertise, and level of their mutual interest. Participative leadership is represented by a fully connected network, while Leader–Member eXchange (LMX) is represented by a fully connected network of in-group members and several out-group members connected only to the leader. Our simulation results show that, in general, participative leadership promotes mental model convergence better than LMX. In the meantime, the team performance improvement is achieved by participative leadership only when members have both heterogeneous domains of expertise and strong mutual interest. In all other conditions, participative leadership causes the worst degradation of team performance through team development processes, while LMX is the best to minimize such team degradation. Implications and suggestions for future research are provided.     

Information systems,inter-functional collaboration and innovation in Taiwanese high-tech manufacturing firms

This article investigates whether (1) cross-functional integration within a firm and the use of information systems (IS) that support information sharing with external parties can enhance integration across the supply chain and wider networks and (2) whether collaboration with customers, suppliers and other external parties leads to increased supply chain performance in terms of new product development and introduction of new processes. Data from a high-quality survey carried out in Taiwan in 2009 were used, and appropriate econometric models were applied. Results show that the adoption of IS that enhance information sharing is vital not only for the effective communication with suppliers and with wider network members, but their adoption also has a direct effect across a firm's innovative effort. Cross-functional integration appears to matter only for the introduction of an innovative process. Collaboration with customers and suppliers affected a product's design and its overall features and functionality, respectively.