A maximal inequality for continuous martingales and \(M\)-estimation in a Gaussian white noise model (Q1970485)

In a Gaussian white noise model the maximizer of a special functional of the unknown regression function is to be estimated. This is achieved by adaptive techniques which are well elaborated, e.g., in empirical process theory. To obtain proper convergence results, the estimator needs to be represented as the argmax of a properly scaled stochastic process which converges in distribution.