cdf_laplace.m | Evaluates the standard Laplace cdf. |

CE_example_gibbs.m | Estimates the shortest path in the bridge example using importance sampling, by first simulating from the optimal importance sampling density via Gibbs sampling, and then using the corresponding CE-optimal estimates as parameters. Uses h.m. |

compsum.m | Estimates an overflow probability associated with a random (compound) sum, arising from a problem in risk insurance, using the optimal exponential twist. |

h.m | Returns the shortest path for the bridge example. |

loss_probab.m | Determines the probability of large portfolio losses in a certain financial model, by first simulating from the optimal importance sampling density via the hit-and-run algorithm, and then using the CE-optimal estimates for importance sampling. Uses score.m. |

NeyPea.m | Estimating an overflow probability associated with a Neyman-Pearson test using the optimal state-independent change of measure. |

OU_process_splitting.m | Estimates a hitting probability of a two-dimensional Ornstein-Uhlenbeck process, using splitting. Uses ou_split.m |

ou_split.m | Implements exact sampling from a two-dimensional Ornstein-Uhlenbeck process, for use with the hitting probability example. |

polkinex.m | Estimates the tail probability of the steady-state waiting time in an M/G/1 queue, using a control variable estimator suggested by the Pollaczek-Khinchin formula. |

score.m | Implements a portfolio loss function. |

siegmund.m | Estimates the hitting time of a positive barrier for a random walk with negative drift via Siegmund's algorithm. |

state_dependent_IS_Laplace.m | Estimates the overflow probability of a sum of iid Laplace terms, using a state-dependent importance sampling scheme. Uses cdf_laplace.m. |

sumlognor.m | Estimates an overflow probability of the sum of independent non-identically distributed log-normals, a problem arising in the computation of certain portfolio measures, using conditioning. |

waitGG1.m | Estimates the steady-state waiting time in an M/M/1 queue via importance sampling with the optimal state-independent change of measure. |

back