sanovprob

Computes the classical Sanov upper bound:

Variational flow-based methods for modeling rare events using
Kullback–Leibler (KL) divergence, normalizing flows, Girsanov change of
measure, and Freidlin–Wentzell action functionals. The package provides
tools for rare-event inference, minimum-action paths, and quasi-potential
computation in stochastic dynamical systems. Methods are based on
Rezende and Mohamed (2015) <doi:10.48550/arXiv.1505.05770>,
Girsanov (1960) <doi:10.1137/1105027>,
and Freidlin and Wentzell (2012, ISBN:978-0387955477).

Pietro Piu

rareflow

Variational Flow-Based Inference for Rare Events and Large
Deviations

sanovprob function

<dl><dt>Q</dt>
<dd>Observed empirical distribution.</dd>
<dt>P</dt>
<dd>True distribution.</dd>
<dt>n</dt>
<dd>Sample size.</dd></dl>

Arguments

Sanov Probability Bound — sanovprob

<dl>

<dt>Q</dt>
<dd>Observed empirical distribution.</dd>


<dt>P</dt>
<dd>True distribution.</dd>


<dt>n</dt>
<dd>Sample size.</dd>

</dl>

sanovprob: Sanov Probability Bound

Description

Usage

Value

Arguments

Details