pystra.active_learning.importance

Use it: User guide · Worked example · Theory

Active importance sampling with explicit Gaussian mixture proposals.

Design-point AK-IS: Echard et al. (2013), doi:10.1016/j.ress.2012.10.008. This variant supports multiple supplied centers and a defensive normal component. It does not automatically discover failure modes or apply a true-model correction.

Classes

ImportanceSamplingDiagnostics	IID likelihood-weight diagnostics conditional on a frozen surrogate.
ImportanceSamplingEstimator	Independent Gaussian-mixture importance sampling for active reliability.