Online b-matching with stochastic rewards has optimal competitive ratio 1-1/e: impossible to beat against the stochastic benchmark, and achieved by StochasticBalance as capacities grow.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.DS 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Online $b$-Matching with Stochastic Rewards
Online b-matching with stochastic rewards has optimal competitive ratio 1-1/e: impossible to beat against the stochastic benchmark, and achieved by StochasticBalance as capacities grow.