We develop efficient importance sampling estimators of certain rare event probabilities involving compound point processes. Our approach is based on the state-dependent techniques developed in (Dupuis & Wang 2004) and subsequent work. The design of the estimators departs from past literature to accommodate the point process setting. Namely, the state-dependent change of measure is updated not at event arrivals but over a deterministic time grid. Several common criteria for the optimality of the estimators are analyzed. Numerical results illustrate the advantages of the proposed estimators in an application setting.
- Start date: 2016-04-05 11:00:00
- End date: 2016-04-05 12:30:00
- Venue: 639 Evans Hall at UC Berkeley
- Address: 639 Evans Hall, Berkeley, CA, 94720