Title :
A simple sequential stopping rule for Monte Carlo Simulation
Author :
Mendo, Luis ; Hernando, Jose M.
Author_Institution :
Signals, Syst. & Radiocommun. Dept., Polytech. Univ., Madrid, Spain
Abstract :
In this paper, a sequential stopping rule for the estimation of a probability p by means of Monte Carlo simulation is analyzed. It is shown that the proposed estimator is almost unbiased, and guarantees a given relative precision irrespective of p. Under very mild conditions, the method also guarantees a certain confidence level for a given relative estimation error, provided that p does not exceed a maximum value. An extension to importance sampling is discussed.
Keywords :
Monte Carlo methods; estimation theory; probability; Monte Carlo simulation; estimation error; probability estimation; simple sequential stopping rule; Communications Society; Discrete event simulation; Estimation error; Monte Carlo methods; Parameter estimation; Power system reliability; Sampling methods; Size measurement; State estimation; Sufficient conditions; Importance sampling (IS); Monte Carlo (MC) methods; sequential stopping rule; simulation;
Journal_Title :
Communications, IEEE Transactions on
DOI :
10.1109/TCOMM.2005.863780
