Frank R. Libsch, S.C. Lien
IBM J. Res. Dev
This paper discusses the application of the likelihood ratio gradient estimator to simulations of large Markovian models of highly dependable systems. Extensive empirical work, as well as some mathematical analysis of small dependability models, suggests that (in this model setting) the gradient estimators are not significantly more noisy than the estimates of the performance measures themselves. The paper also discusses implementation issues associated with likelihood ratio gradient estimation, as well as some theoretical complements associated with application of the technique to continuous-time Markov chains.
Frank R. Libsch, S.C. Lien
IBM J. Res. Dev
Renu Tewari, Richard P. King, et al.
IS&T/SPIE Electronic Imaging 1996
Hang-Yip Liu, Steffen Schulze, et al.
Proceedings of SPIE - The International Society for Optical Engineering
Nanda Kambhatla
ACL 2004