ON MINIMUM BIAS AND VARIANCE ESTIMATION FOR PARAMETRIC
MODELS WITH SHRINKING CONTAMINATION
Abstract: A close relationship is derived between optimal -estimation and optimal robust
testing for shrinking contaminations. Explicit formulas are given for solutions when the loss
is denned as convex combination of asymptotic maximum bias span and variance.
Neighbourhoods are described in terms of a general class of special capacities.
2000 AMS Mathematics Subject Classification: Primary: -; Secondary: -;
Key words and phrases: -