JMLR
2006
13 years 11 months ago
2006
Abstract. We give dimension-free and data-dependent bounds for linear multi-task learning where a common linear operator is chosen to preprocess data for a vector of task speci...c...