Latent Preference Bandits

Newton Mwai, Emil Carlsson, Fredrik D. Johansson

Bandit algorithms are guaranteed to solve diverse sequential decision-making problems, provided that a sufficient exploration budget is available. However, learning from scratch is…