Hybrid reward architecture for reinforcement learning

Hybrid reward architecture for reinforcement learning

Aspects provided herein are relevant to machine learning techniques, including decomposing single-agent reinforcement learning problems into simpler problems addressed by multiple agents. Actions proposed by the multiple agents are then aggregated using an aggregator, which selects an action to take...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Van Seijen, Harm Hendrik, Fatemi Booshehri, Seyed Mehdi, Laroche, Romain Michel Henri, Romoff, Joshua Samuel
Format: Patent
Sprache:eng
Schlagworte:
CALCULATING
COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
COMPUTING
COUNTING
PHYSICS
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Aspects provided herein are relevant to machine learning techniques, including decomposing single-agent reinforcement learning problems into simpler problems addressed by multiple agents. Actions proposed by the multiple agents are then aggregated using an aggregator, which selects an action to take with respect to an environment. Aspects provided herein are also relevant to a hybrid reward model.