英和和英

名詞

sparsistency (uncountable)

1. (statistics) Let ${\displaystyle \mathbf {b} }$ be a vector and define the support ${\displaystyle \operatorname {supp} (\mathbf {b} )=\{i:\mathbf {b} _{i}\neq 0\}}$ where ${\displaystyle \mathbf {b} _{i}}$ is the ${\displaystyle i}$th element of ${\displaystyle \mathbf {b} }$. Let ${\displaystyle {\hat {\mathbf {b} }}}$ be an estimator for ${\displaystyle \mathbf {b} }$. Then sparsistency is the property that the support of the estimator converges to the true support as the number of samples grows to infinity.
   • 2015, Edouard Ollier, Vivian Viallon, “Regression modeling on stratified data: automatic and covariate-specific selection of the reference stratum with simple ${\displaystyle L_{1}}$-norm penalties”, in arXiv‎^[1]:

     Its implementation can be done with available packages under a variety of models and, in the linear regression model, we show it is sparsistent under conditions similar to those ensuring sparsistency for an oracular version of the reference stratum strategy.