Error Exponent
A core metric for measuring the performance of best-arm identification algorithms. Based on the large deviation theory framework, it describes the rate at which the algorithm's error probability decays exponentially as the budget increases.
Timeline (last 90 days)
For Gaussian arms with a common variance, SR's worst-case error exponent (after normalization by H₁(P)) is already optimal up to a constant factor.
The error exponent is the core metric for measuring the performance of best-arm identification algorithms: under the large deviation theory framework, as the total budget tends to infinity, the algorithm's probability of error decays at an exponential rate, and that decay rate is the error exponent.
All Facts (2)
For Gaussian arms with a common variance, SR's worst-case error exponent (after normalization by H₁(P)) is already optimal up to a constant factor.
50%UnverifiedThe error exponent is the core metric for measuring the performance of best-arm identification algorithms: under the large deviation theory framework, as the total budget tends to infinity, the algorithm's probability of error decays at an exponential rate, and that decay rate is the error exponent.
50%