We find two fundamental statistics in discriminant analysis on which many influence measures depend. A few other measures are also proposed and discussed. Approximate contours of the measure can be plotted to reveal influence information.

G. R. Ducharme and Y. Lepage (1986, J. Roy. Statist. Soc. Ser. B48, 197-205) presented the strong collapsibility of odds ratio in 22K tables. However, the concept is not suitable for an ordinal background variable since it is meaningless to pool nonadjacent levels in this case. In this paper, we...

Count data with excess zeros arises in many contexts. Here our concern is to develop a Bayesian analysis for the zero-inflated generalized Poisson (ZIGP) regression model to address this problem. This model provides a useful generalization of zero-inflated Poisson model since the generalized...

Least absolute deviations (LAD) regression, i.e. L1 regression, is more resistant to the outliers in the response variable than the least-squares regression, but is relatively sensitive to outlying observations in explanatory variables. However, some but few attentions have been contributed to...

We consider a new mathematical object called preferred point [alpha]-manifold, which is a natural extension of Lauritzen (1987) and Critchley et al. (1993). A sufficient and necessary condition is given to guarantee that preferred point [alpha]-geometry subsumes a statistical manifold. We...