TITLE:  Partial Mixture Estimation and Outlier
        Detection in Data and Regression

ABSTRACT:

A key step in the formulation of statistical models
is the likelihood function.  In many practical situations,
a reasonable model is known for a subset of the data.
We explore the idea of robustly fitting a mixture of multivariate
Gaussian densities, but when the number of components
is intentionally too few.  Using a minimum distance
criterion, we show how useful results may be
obtained in practice.  Application areas are numerous,
and examples will be provided.  We will reexamine
the classical Boston housing data, with spatial
views of the residuals, as well as similar California data.