Preference Elicitation within Framework of Fully Probabilistic Design of Decision Strategies

Design and Self-Assembling Behaviour of Calamitic Reactive Mesogens with Lateral Methyl and Methoxy Substituents and Vinyl Terminal Group

“…The self-organisation behaviour build-up on the intermolecular interactions can be effectively tuned by the appropriate molecular design, e.g., by constructing the macromolecular system from the mesogenic units with the reactive terminal groups [ 15 , 16 , 17 ]. However, it is very difficult to theoretically predict, describe and keep under control the macroscopic parameters of the self-assembling materials [ 18 , 19 , 20 ]. Many experimental efforts in reactive mesogen design [ 4 ] have been carried out while building-up new photosensitive and photo-controllable macromolecular materials [ 10 ], incorporating a photo-responsive azo group into the molecular core.…”

On the crowdsourcing of behaviors for autonomous agents

“…In [5], a two-sided data marketplace is formalized and management algorithms are introduced, while techniques allowing robots with the same task to re-use models have been introduced in [6]. We also recall works on multi-agent reinforcement learning [7], where agents collaborate in order to learn a common task, and on preference elicitation within a probabilistic framework of decision-making [8], where an individual makes optimal decisions based on his/her own preferences. As we shall see, crowdsourcing can be formalized as a data-driven control problem [9]- [11] and recent results include G. Russo is with the Dept.…”