4 papers
On Generalizing Preference-based Reinforcement Learning to Multi-Objective Problems
Simone Drago, Marco Mussi, Leonardo Bianconi +1
We study reinforcement learning (RL) from pairwise human-generated trajectory comparisons. We generalize preference-based RL by formalizing a novel setting in which the expert can…
Generalized Kernelized Bandits: A Novel Self-Normalized Bernstein-Like Dimension-Free Inequality and Regret Bounds
Alberto Maria Metelli, Simone Drago, Marco Mussi
We study the regret minimization problem in the novel setting of generalized kernelized bandits (GKBs), where we optimize an unknown function belonging to a reproducing kerne…
A Refined Analysis of UCBVI
Simone Drago, Marco Mussi, Alberto Maria Metelli
In this work, we provide a refined analysis of the UCBVI algorithm (Azar et al., 2017), improving both the bonus terms and the regret analysis. Additionally, we compare our version…
Open Problem: Tight Bounds for Kernelized Multi-Armed Bandits with Bernoulli Rewards
Marco Mussi, Simone Drago, Alberto Maria Metelli
We consider Kernelized Bandits (KBs) to optimize a function belonging to the Reproducing Kernel Hilbert Space (RKHS) . Mainstream…