An optimal algorithm for Monte Carlo estimation

Author

Dagum, Paul ; Karp, Richard ; Luby, Michael ; Ross, Sheldon

Author_Institution

Sect. on Med. Inf., Stanford Univ. Sch. of Med., Palo Alto, CA, USA

fYear

1995

fDate

23-25 Oct 1995

Firstpage

142

Lastpage

149

Abstract

A typical approach to estimate an unknown quantity μ is to design an experiment that produces a random variable Z distributed in [O,1] with E[Z]=μ, run this experiment independently a number of times and use the average of the outcomes as the estimate. In this paper, we consider the case when no a priori information about Z is known except that is distributed in [0,1]. We describe an approximation algorithm AA which, given ε and δ, when running independent experiments with respect to any Z, produces an estimate that is within a factor 1+ε of μ with probability at least 1-δ. We prove that the expected number of experiments ran by AA (which depends on Z) is optimal to within a constant factor for every Z

Keywords

Monte Carlo methods; parallel algorithms; Monte Carlo estimation; a priori information; approximation algorithm; optimal algorithm; Algorithm design and analysis; Approximation algorithms; Computer science; Design for experiments; Estimation theory; Monte Carlo methods; Probability; Random variables; Silicon compounds; Upper bound;

fLanguage

English

Publisher

ieee

Conference_Titel

Foundations of Computer Science, 1995. Proceedings., 36th Annual Symposium on

Conference_Location

Milwaukee, WI

ISSN

0272-5428

Print_ISBN

0-8186-7183-1

Type

conf

DOI

10.1109/SFCS.1995.492471

Filename

492471