D

DistroSA

Project information

This package develops novel algorithmic approaches for distributional sensitivity analysis. Specifically, we develop a new empirical algorithm for calculating derivative of any probability distribution parameters with respect to perturbations in realization of the random variable. This is critical for machine learning applications that seek to fit a probability distribution to observational data.

sensitivity ... probability ...

15 Commits
2 Branches
1 Tag

README
BSD 3-Clause "New" or "Revised" License

Created on

February 08, 2024

Loading