Pruning Rules for Learning Parsimonious Context Trees

Show full item record



Eggeling , R & Koivisto , M K H 2016 , Pruning Rules for Learning Parsimonious Context Trees . in A Ihler & D Janzing (eds) , Uncertainty in Artificial Intelligence : Proceedings of the Thirty-Second Conference (2016) June 25-29, 2016, Jersey City, New Jersey, USA . AUAI Press , Corvallis, Oregon , pp. 152-161 , Conference on Uncertainty in Artificial Intelligence , Jersey City, New Jersey , United States , 25/06/2016 .

Title: Pruning Rules for Learning Parsimonious Context Trees
Author: Eggeling, Ralf; Koivisto, Mikko Kalle Henrik
Other contributor: Ihler, Alexander
Janzing, Dominik
Contributor organization: Helsinki Institute for Information Technology
Department of Computer Science
Sums of Products research group / Mikko Koivisto
Finnish Centre of Excellence in Algorithmic Data Analysis Research (Algodan)
Publisher: AUAI Press
Date: 2016
Language: eng
Number of pages: 10
Belongs to series: Uncertainty in Artificial Intelligence
ISBN: 978-0-9966431-1-5
Abstract: We give a novel algorithm for finding a parsimonious context tree (PCT) that best fits a given data set. PCTs extend traditional context trees by allowing context-specific grouping of the states of a context variable, also enabling skipping the variable. However, they gain statistical efficiency at the cost of computational efficiency, as the search space of PCTs is of tremendous size. We propose pruning rules based on efficiently computable score upper bounds with the aim of reducing this search space significantly. While our concrete bounds exploit properties of the BIC score, the ideas apply also to other scoring functions. Empirical results show that our algorithm is typically an order-of-magnitude faster than a recently proposed memory-intensive algorithm, or alternatively, about equally fast but using dramatically less memory.
Subject: 113 Computer and information sciences
Peer reviewed: Yes
Usage restriction: openAccess
Self-archived version: publishedVersion

Files in this item

Total number of downloads: Loading...

Files Size Format View
UAI2016_Paper66.pdf 586.0Kb PDF View/Open

This item appears in the following Collection(s)

Show full item record