Alexandra Carpentier et Nicolas Verzelen, Adaptive estimation of the sparsity in the Gaussian vector model, Annals of statistics , 47(1), 2019, pp. 93-126
Consider the Gaussian vector model with mean value .. We study the twin problems of estimating the number ...0 of nonzero components of . and testing whether ...0 is smaller than some value. For testing, we establish the minimax separation distances for this model and introduce a minimax adaptive test. Extensions to the case of unknown variance are also discussed. Rewriting the estimation of ...0 as a multiple testing problem of all hypotheses {...0.q}, we both derive a new way of assessing the optimality of a sparsity estimator and we exhibit such an optimal procedure. This general approach provides a roadmap for estimating the complexity of the signal in various statistical models.