Competing with automata-based expert sequences

Mehryar Mohri, Scott Yang

Research output: Contribution to conferencePaper

Abstract

We consider a general framework of online learning with expert advice where regret is defined with respect to sequences of experts accepted by a weighted automaton. Our framework covers several problems previously studied, including competing against k-shifting experts. We give a series of algorithms for this problem, including an automata-based algorithm extending weighted-majority and more efficient algorithms based on the notion of failure transitions. We further present efficient algorithms based on an approximation of the competitor automaton, in particular n-gram models obtained by minimizing the ∞-Rényi divergence, and present an extensive study of the approximation properties of such models. Finally, we also extend our algorithms and results to the framework of sleeping experts.

Original languageEnglish (US)
Pages1732-1740
Number of pages9
StatePublished - Jan 1 2018
Event21st International Conference on Artificial Intelligence and Statistics, AISTATS 2018 - Playa Blanca, Lanzarote, Canary Islands, Spain
Duration: Apr 9 2018Apr 11 2018

Conference

Conference21st International Conference on Artificial Intelligence and Statistics, AISTATS 2018
CountrySpain
CityPlaya Blanca, Lanzarote, Canary Islands
Period4/9/184/11/18

Fingerprint

Automata
Efficient Algorithms
Weighted Automata
N-gram
Regret
Online Learning
Approximation Property
Divergence
Cover
Series
Approximation
Model
Framework

ASJC Scopus subject areas

  • Statistics and Probability
  • Artificial Intelligence

Cite this

Mohri, M., & Yang, S. (2018). Competing with automata-based expert sequences. 1732-1740. Paper presented at 21st International Conference on Artificial Intelligence and Statistics, AISTATS 2018, Playa Blanca, Lanzarote, Canary Islands, Spain.

Competing with automata-based expert sequences. / Mohri, Mehryar; Yang, Scott.

2018. 1732-1740 Paper presented at 21st International Conference on Artificial Intelligence and Statistics, AISTATS 2018, Playa Blanca, Lanzarote, Canary Islands, Spain.

Research output: Contribution to conferencePaper

Mohri, M & Yang, S 2018, 'Competing with automata-based expert sequences' Paper presented at 21st International Conference on Artificial Intelligence and Statistics, AISTATS 2018, Playa Blanca, Lanzarote, Canary Islands, Spain, 4/9/18 - 4/11/18, pp. 1732-1740.
Mohri M, Yang S. Competing with automata-based expert sequences. 2018. Paper presented at 21st International Conference on Artificial Intelligence and Statistics, AISTATS 2018, Playa Blanca, Lanzarote, Canary Islands, Spain.
Mohri, Mehryar ; Yang, Scott. / Competing with automata-based expert sequences. Paper presented at 21st International Conference on Artificial Intelligence and Statistics, AISTATS 2018, Playa Blanca, Lanzarote, Canary Islands, Spain.9 p.
@conference{10b1d3971ecc4da2b7e69b06f74c3ad5,
title = "Competing with automata-based expert sequences",
abstract = "We consider a general framework of online learning with expert advice where regret is defined with respect to sequences of experts accepted by a weighted automaton. Our framework covers several problems previously studied, including competing against k-shifting experts. We give a series of algorithms for this problem, including an automata-based algorithm extending weighted-majority and more efficient algorithms based on the notion of failure transitions. We further present efficient algorithms based on an approximation of the competitor automaton, in particular n-gram models obtained by minimizing the ∞-R{\'e}nyi divergence, and present an extensive study of the approximation properties of such models. Finally, we also extend our algorithms and results to the framework of sleeping experts.",
author = "Mehryar Mohri and Scott Yang",
year = "2018",
month = "1",
day = "1",
language = "English (US)",
pages = "1732--1740",
note = "21st International Conference on Artificial Intelligence and Statistics, AISTATS 2018 ; Conference date: 09-04-2018 Through 11-04-2018",

}

TY - CONF

T1 - Competing with automata-based expert sequences

AU - Mohri, Mehryar

AU - Yang, Scott

PY - 2018/1/1

Y1 - 2018/1/1

N2 - We consider a general framework of online learning with expert advice where regret is defined with respect to sequences of experts accepted by a weighted automaton. Our framework covers several problems previously studied, including competing against k-shifting experts. We give a series of algorithms for this problem, including an automata-based algorithm extending weighted-majority and more efficient algorithms based on the notion of failure transitions. We further present efficient algorithms based on an approximation of the competitor automaton, in particular n-gram models obtained by minimizing the ∞-Rényi divergence, and present an extensive study of the approximation properties of such models. Finally, we also extend our algorithms and results to the framework of sleeping experts.

AB - We consider a general framework of online learning with expert advice where regret is defined with respect to sequences of experts accepted by a weighted automaton. Our framework covers several problems previously studied, including competing against k-shifting experts. We give a series of algorithms for this problem, including an automata-based algorithm extending weighted-majority and more efficient algorithms based on the notion of failure transitions. We further present efficient algorithms based on an approximation of the competitor automaton, in particular n-gram models obtained by minimizing the ∞-Rényi divergence, and present an extensive study of the approximation properties of such models. Finally, we also extend our algorithms and results to the framework of sleeping experts.

UR - http://www.scopus.com/inward/record.url?scp=85051140024&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85051140024&partnerID=8YFLogxK

M3 - Paper

SP - 1732

EP - 1740

ER -