This paper deals with aggregation of estimators in the context of regression with fixed design, with heteroscedastic and subgaussian noise. We relate the task of aggregating a finite family of affine estimators to the concentration of quadratic forms of the noise vector, and we derive sharp...

Discriminant analysis for two data sets in IRd with probability densities f and g can be based on the estimation of the set G = {x : f(x) Ï g(x)}. We consider applications where it is appropriate to assume that the region G has a smooth boundary. In particular, this assumption makes sense if...

Discriminant analysis for two data sets in IRd with probability densities f and g can be based on the estimation of the set G = {x : f(x) Ï g(x)}. We consider applications where it is appropriate to assume that the region G has a smooth boundary. In particular, this assumption makes sense if...