David Schnoerr |

Tue 04 Jul 2017, 11:00 - 12:00 |

IF 4.31/4.33 |

If you have a question about this talk, please contact: Gareth Beedham (gbeedham)

*Joint work with Botond Cseke, Ramon Grima and Guido Sanguinetti.*

**Approximating first-passage time distributions via sequential Bayesian computation **

*Many systems in nature consist of stochastically interacting agents or particles. Stochastic processes have been widely used to model such systems, yet they are notoriously difficult to analyse. We consider the problem of computing first-passage times for Markov jump processes, that is, the time it takes a process to first cross a certain threshold. I will show that this important class of generally intractable problems can be exactly recast in terms of a Bayesian inference problem by introducing auxiliary observations. This leads us to derive an efficient approximation scheme to compute first-passage time distributions by solving a small, closed set of ordinary differential equations. **The method is accurate and orders of magnitude faster than existing approaches, enabling hitherto computationally prohibitive tasks such as sensitivity analysis. We apply it to an epidemic model, an entrained oscillator and a trimerisation process, and show good agreement with exact stochastic simulations. *