MAUPG, a scenario-conditioned MADDPG with hindsight replay, generalizes one cooperative VIP-protection task across four simulator scenarios and beats scenario-specific policies.
Rusu, Joel Veness, Marc G
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.MA 1years
2019 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Universal Policies to Learn Them All
MAUPG, a scenario-conditioned MADDPG with hindsight replay, generalizes one cooperative VIP-protection task across four simulator scenarios and beats scenario-specific policies.