Reformulating Inference Problems through Selective Conditioning

From Tetherless World Wiki

Jump to: navigation, search

Citation: Paul Dagum and Eric Horvitz. (1992) Reformulating Inference Problems through Selective Conditioning. In KSL-92-50, 1992.

Publication techreport ( Edit )
type Technical Report
bibtype techreport
Bibtex basics
author Paul Dagum and Eric Horvitz
title Reformulating Inference Problems through Selective Conditioning
number KSL-92-50
institution Knowledge Systems, AI Laboratory
address Stanford, CA
year 1992
Bibtex more
Access Paper
abstract We describe how we selectively reformulate portions of a belief network that pose difficulties for solution with a stochastic-simulation algorithm. We employ the selective conditioning approach to target specific nodes in a belief network for decomposition, based on the contribution the nodes make to the tractability of stochastic simulation. We review previous work on BNRAS algorithms-randomized approximation algorithms for probabilistic inference. We show how selective conditioning can be employed to reformulate a single BNRAS problem into multiple tractable BNRAS simulation problems. We discuss how we can use another simulation algorithm-logic sampling-to solve a component of the inference problem that provides a means for knitting the solutions of individual subproblems into a final result. Finally, we analyze tradeoffs among the computational subtasks associated with the selective conditioning approach to reformulation.

KSL Technical Report ID: KSL-92-50
Facts about Reformulating Inference Problems through Selective ConditioningRDF feed
Abstract We describe how we selectively reformulate … We describe how we selectively reformulate portions of a belief network that pose difficulties for solution with a stochastic-simulation algorithm. We employ the selective conditioning approach to target specific nodes in a belief network for decomposition, based on the contribution the nodes make to the tractability of stochastic simulation. We review previous work on BNRAS algorithms-randomized approximation algorithms for probabilistic inference. We show how selective conditioning can be employed to reformulate a single BNRAS problem into multiple tractable BNRAS simulation problems. We discuss how we can use another simulation algorithm-logic sampling-to solve a component of the inference problem that provides a means for knitting the solutions of individual subproblems into a final result. Finally, we analyze tradeoffs among the computational subtasks associated with the selective conditioning approach to reformulation. ve conditioning approach to reformulation.
Address Stanford, CA  +
Author Paul Dagum and Eric Horvitz  +
Bibtype techreport  +
Has author Paul Dagum and Eric Horvitz  +
Has identifier KSL-92-50  +
Has publishing details 1992  +
Has title Reformulating Inference Problems through Selective Conditioning  +
Has where published KSL-92-50  +
Has year 1992  +
Institution Knowledge Systems, AI Laboratory  +
Ksl tr id KSL-92-50  +
Number KSL-92-50  +
Process note YES  +
Title Reformulating Inference Problems through Selective Conditioning  +
Year 1992  +
Retrieved from "http://tw.rpi.edu/wiki/Reformulating_Inference_Problems_through_Selective_Conditioning"
Personal tools