1 paper · 1 filter
Fabian Obster, Christian Heumann
In the presence of grouped covariates, we propose a framework for boosting that allows to enforce sparsity within and between groups. By using component-wise and group-wise gradien…