Fixed-Confidence, Fixed-Tolerance Guarantees for Ranking-and-Selection Procedures
نویسندگان
چکیده
Ever since the conception of statistical ranking-and-selection (R8S) problem, a predominant approach has been indifference-zone (IZ) formulation. Under IZ formulation, R8S procedures are designed to provide guarantee on probability correct selection (PCS) whenever performance best system exceeds that second-best by specified amount. We discuss shortcomings this and argue providing good (PGS)—selecting whose is within tolerance best—is more justifiable goal. Unfortunately, form fixed-confidence, fixed-tolerance received far less attention simulation community. present an overview PGS with aim reorienting community toward examine numerous techniques for proving guarantee, including sufficient conditions under which subset-selection deliver IZ-inspired PCS also guarantee.
منابع مشابه
Structured Best Arm Identification with Fixed Confidence
We study the problem of identifying the best action among a set of possible options when the value of each action is given by a mapping from a number of noisy micro-observables in the so-called fixed confidence setting. Our main motivation is the application to the minimax game search, which has been a major topic of interest in artificial intelligence. In this paper we introduce an abstract se...
متن کاملConfidence Intervals with Fixed Proportional Accuracy
The problem of setting a confidence interval with fixed proportional accuracy for the mean of a normal distribution is considered. A sequential procedure is proposed; and asymptotic expansions for its average coverage probabilities and its expected sample size are obtained. The procedure is shown to have asymptotically minimal expected sample size, subject to a constraint on the error probabili...
متن کاملOptimal Best Arm Identification with Fixed Confidence
We give a complete characterization of the complexity of best-arm identification in one-parameter bandit problems. We prove a new, tight lower bound on the sample complexity. We propose the ‘Track-and-Stop’ strategy, which we prove to be asymptotically optimal. It consists in a new sampling rule (which tracks the optimal proportions of arm draws highlighted by the lower bound) and in a stopping...
متن کاملOptimal Odd Arm Identification with Fixed Confidence
The problem of detecting an odd arm from a set of K arms of a multi-armed bandit, with fixed confidence, is studied in a sequential decision-making scenario. Each arm’s signal follows a distribution from a vector exponential family. All arms have the same parameters except the odd arm. The actual parameters of the odd and non-odd arms are unknown to the decision maker. Further, the decision mak...
متن کاملProbabilistic QoS Guarantees: Fixed and Dynamic Priority Scheduling
Generally, with a limited number of fixed priorities ( ) within a network, applications sharing a same priority are scheduled according to First in First out ( I O) discipline. This policy is called FP/FIFO. However real time applications have deadline requirements which aren’t taken into account with the I O discipline. That is why we propose to introduce the Earliest Deadline First ( D ) disc...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: ACM Transactions on Modeling and Computer Simulation
سال: 2021
ISSN: ['1049-3301', '1558-1195']
DOI: https://doi.org/10.1145/3432754