Estimating Time Averages via Randomly-Spaced Observations
The paper deals with continuous-time averages of discrete-event systems via randomly-spaced observations: The aim is to estimate the limit, including the construction of confidence intervals, \(r=\lim r(t)\), a.s. as \(t\to \infty\), where \(r(t)=(1/t)\int^{t}_{0}f(X(s))ds\), f is a real-valued function, \(X(t)=\sum \tilde X_ kI_{[T_ k,T_{k+1})}(t)\), \(I_ A(t)\) is the indicator of A, \(0=T_ 0<T_ 1<..\). are event times and \(\tilde X_ k\in {\mathbb{R}}^ d\) is a state at time \(T_ k.\) The authors develop a point-process framework and use it to generalize both regenerative and stationary-process oriented simulation methodologies. Consistency and central limit theorems for the time averages are proved and an effective bias-reducing jackknife is suggested.
- Estimation Methods for Delays in Non-regenerative Discrete-Event Systems
- Event and time averages: a review
- Correlation-inducing variance reduction in regenerative simulation
- The average number of events per time unit and its estimation
- Estimating the steady-state mean from short transient simulations
This page was built for publication: Estimating Time Averages via Randomly-Spaced Observations
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3750863)