English

Moment-Matching Conditions for Exponential Families with Conditioning or Hidden Data

Machine Learning 2020-01-28 v1 Machine Learning

Abstract

Maximum likelihood learning with exponential families leads to moment-matching of the sufficient statistics, a classic result. This can be generalized to conditional exponential families and/or when there are hidden data. This document gives a first-principles explanation of these generalized moment-matching conditions, along with a self-contained derivation.

Keywords

Cite

@article{arxiv.2001.09771,
  title  = {Moment-Matching Conditions for Exponential Families with Conditioning or Hidden Data},
  author = {Justin Domke},
  journal= {arXiv preprint arXiv:2001.09771},
  year   = {2020}
}