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 ...

Created on

February 08, 2024

All
Wiki
Team

Loading