1 paper
Steven Bilaj, Sofien Dhouib, Setareh Maghsudi
We study the problem of meta-learning several contextual stochastic bandits tasks by leveraging their concentration around a low-dimensional affine subspace, which we learn via onl…