Bandits on graphs and structures

We investigate the structural properties of certain sequential decision-making problems with limited feedback (bandits) in order to bring the known algorithmic solutions closer to a practical use. In the first part, we put a special emphasis on structures that can be represented as graphs on actions, in the second part we study the large action spaces that can be of exponential size in the number of base actions or even infinite. We show how to take advantage of structures over the actions and (provably) learn faster