Active Preference Learning using Maximum Regret

Nils Wilde, Dana Kulic, Stephen L. Smith

We study active preference learning as a framework for intuitively specifying the behaviour of autonomous robots. In active preference learning, a user chooses the preferred behavi…