We focus on neuro-dynamic programming methods to learn state-action value functions and outline some of the inherent problems to be faced, when performing reinforcement learning in...
We investigate the problem of non-covariant behavior of policy gradient reinforcement learning algorithms. The policy gradient approach is amenable to analysis by information geom...
Policy-based management (PBM) aims to provide flexibility in the management of resources so as to readily reflect changing business goals. However, as organizations increasingly u...
Modern multiuser software systems have adopted RoleBased Access Control (RBAC) for authorization management. This paper presents a formal model for RBAC policy validation and a st...
Marco Pistoia, Stephen J. Fink, Robert J. Flynn, E...