An α-approximate portfolio of RL policies can cover all p-mean social welfare objectives for p ≤ 1 with size O(log κ / log(1/α)), and the paper gives algorithms and experiments for this.
In our experiments, we set N = 59 and construct |Π| = 200 balancing policies as feasible set
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Navigating the Social Welfare Frontier: Portfolios for Multi-objective Reinforcement Learning
An α-approximate portfolio of RL policies can cover all p-mean social welfare objectives for p ≤ 1 with size O(log κ / log(1/α)), and the paper gives algorithms and experiments for this.