Discovering hidden variables in noisy-or networks using quartet tests

Yacine Jernite, Yoni Halpern, David Sontag

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We give a polynomial-time algorithm for provably learning the structure and parameters of bipartite noisy-or Bayesian networks of binary variables where the top layer is completely hidden. Unsupervised learning of these models is a form of discrete factor analysis, enabling the discovery of hidden variables and their causal relationships with observed data. We obtain an efficient learning algorithm for a family of Bayesian networks that we call quartet-learnable. For each latent variable, the existence of a singly-coupled quartet allows us to uniquely identify and learn all parameters involving that latent variable. We give a proof of the polynomial sample complexity of our learning algorithm, and experimentally compare it to variational EM.

Original languageEnglish (US)
Title of host publicationAdvances in Neural Information Processing Systems
PublisherNeural information processing systems foundation
StatePublished - 2013
Event27th Annual Conference on Neural Information Processing Systems, NIPS 2013 - Lake Tahoe, NV, United States
Duration: Dec 5 2013Dec 10 2013

Other

Other27th Annual Conference on Neural Information Processing Systems, NIPS 2013
CountryUnited States
CityLake Tahoe, NV
Period12/5/1312/10/13

Fingerprint

Bayesian networks
Learning algorithms
Polynomials
Unsupervised learning
Factor analysis

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Information Systems
  • Signal Processing

Cite this

Jernite, Y., Halpern, Y., & Sontag, D. (2013). Discovering hidden variables in noisy-or networks using quartet tests. In Advances in Neural Information Processing Systems Neural information processing systems foundation.

Discovering hidden variables in noisy-or networks using quartet tests. / Jernite, Yacine; Halpern, Yoni; Sontag, David.

Advances in Neural Information Processing Systems. Neural information processing systems foundation, 2013.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Jernite, Y, Halpern, Y & Sontag, D 2013, Discovering hidden variables in noisy-or networks using quartet tests. in Advances in Neural Information Processing Systems. Neural information processing systems foundation, 27th Annual Conference on Neural Information Processing Systems, NIPS 2013, Lake Tahoe, NV, United States, 12/5/13.
Jernite Y, Halpern Y, Sontag D. Discovering hidden variables in noisy-or networks using quartet tests. In Advances in Neural Information Processing Systems. Neural information processing systems foundation. 2013
Jernite, Yacine ; Halpern, Yoni ; Sontag, David. / Discovering hidden variables in noisy-or networks using quartet tests. Advances in Neural Information Processing Systems. Neural information processing systems foundation, 2013.
@inproceedings{eab835b032dd4ca48e76b8a137bad22f,
title = "Discovering hidden variables in noisy-or networks using quartet tests",
abstract = "We give a polynomial-time algorithm for provably learning the structure and parameters of bipartite noisy-or Bayesian networks of binary variables where the top layer is completely hidden. Unsupervised learning of these models is a form of discrete factor analysis, enabling the discovery of hidden variables and their causal relationships with observed data. We obtain an efficient learning algorithm for a family of Bayesian networks that we call quartet-learnable. For each latent variable, the existence of a singly-coupled quartet allows us to uniquely identify and learn all parameters involving that latent variable. We give a proof of the polynomial sample complexity of our learning algorithm, and experimentally compare it to variational EM.",
author = "Yacine Jernite and Yoni Halpern and David Sontag",
year = "2013",
language = "English (US)",
booktitle = "Advances in Neural Information Processing Systems",
publisher = "Neural information processing systems foundation",

}

TY - GEN

T1 - Discovering hidden variables in noisy-or networks using quartet tests

AU - Jernite, Yacine

AU - Halpern, Yoni

AU - Sontag, David

PY - 2013

Y1 - 2013

N2 - We give a polynomial-time algorithm for provably learning the structure and parameters of bipartite noisy-or Bayesian networks of binary variables where the top layer is completely hidden. Unsupervised learning of these models is a form of discrete factor analysis, enabling the discovery of hidden variables and their causal relationships with observed data. We obtain an efficient learning algorithm for a family of Bayesian networks that we call quartet-learnable. For each latent variable, the existence of a singly-coupled quartet allows us to uniquely identify and learn all parameters involving that latent variable. We give a proof of the polynomial sample complexity of our learning algorithm, and experimentally compare it to variational EM.

AB - We give a polynomial-time algorithm for provably learning the structure and parameters of bipartite noisy-or Bayesian networks of binary variables where the top layer is completely hidden. Unsupervised learning of these models is a form of discrete factor analysis, enabling the discovery of hidden variables and their causal relationships with observed data. We obtain an efficient learning algorithm for a family of Bayesian networks that we call quartet-learnable. For each latent variable, the existence of a singly-coupled quartet allows us to uniquely identify and learn all parameters involving that latent variable. We give a proof of the polynomial sample complexity of our learning algorithm, and experimentally compare it to variational EM.

UR - http://www.scopus.com/inward/record.url?scp=84898991472&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84898991472&partnerID=8YFLogxK

M3 - Conference contribution

BT - Advances in Neural Information Processing Systems

PB - Neural information processing systems foundation

ER -