Sparsity-Agnostic Lasso Bandit

                   Min-hwan Oh 1 Garud Iyengar 2 Assaf Zeevi 2

             Abstract                 the traditional MAB problem, here pulling any one arm
                                  provides some information about the unknown parameter
   We consider a stochastic contextual bandit prob-       vector, and hence, insight into the average reward of all the
   lem where the dimension d of the feature vectors       other arms. These contextual bandit algorithms ...