Learning Product Automata

Joshua Moerman

Learning Product Automata

Joshua Moerman

Research output: Chapter in Book/Report/Conference proceeding › Conference Article in proceeding › Academic › peer-review

Abstract

We give an optimisation for active learning algorithms, applicable to learning Moore machines with decomposable outputs. These machines can be decomposed themselves by projecting on each output. This results in smaller components that can then be learnt with fewer queries. We give experimental evidence that this is a useful technique which can reduce the number of queries substantially. Only in some cases the performance is worsened by the slight overhead. Compositional methods are widely used throughout engineering, and the decomposition presented in this article promises to be particularly interesting for learning hardware systems.

Original language	English
Title of host publication	Proceedings of Machine Learning Research
Editors	Olgierd Unold, Witold Dyrka, Wojciech Wieczorek
Publisher	PMLR
Pages	54-66
Number of pages	13
Volume	93
Publication status	Published - 2019
Externally published	Yes
Event	The 14th International Conference on Grammatical Inference - Wrocław, Poland Duration: 5 Sept 2018 → 7 Sept 2018 Conference number: 14 http://icgi2018.pwr.edu.pl/papers

Conference

Conference	The 14th International Conference on Grammatical Inference
Abbreviated title	ICGI2018
Country/Territory	Poland
City	Wrocław
Period	5/09/18 → 7/09/18
Internet address	http://icgi2018.pwr.edu.pl/papers

Access to Document

http://proceedings.mlr.press/v93/moerman19a.html

Learning Product Automata
Moerman, J., May 2017, Cornell University - arXiv, 5 p. (Computing Research Repository).
Research output: Working paper / Preprint › Preprint

Open Access

Cite this

@inproceedings{e83c0b757aa341af905fb67a2ce934d3,

title = "Learning Product Automata",

abstract = "We give an optimisation for active learning algorithms, applicable to learning Moore machines with decomposable outputs. These machines can be decomposed themselves by projecting on each output. This results in smaller components that can then be learnt with fewer queries. We give experimental evidence that this is a useful technique which can reduce the number of queries substantially. Only in some cases the performance is worsened by the slight overhead. Compositional methods are widely used throughout engineering, and the decomposition presented in this article promises to be particularly interesting for learning hardware systems.",

author = "Joshua Moerman",

note = "DBLP License: DBLP's bibliographic metadata records provided through http://dblp.org/ are distributed under a Creative Commons CC0 1.0 Universal Public Domain Dedication. Although the bibliographic metadata records are provided consistent with CC0 1.0 Dedication, the content described by the metadata records is not. Content may be subject to copyright, rights of privacy, rights of publicity and other restrictions.; The 14th International Conference on Grammatical Inference, ICGI2018 ; Conference date: 05-09-2018 Through 07-09-2018",

year = "2019",

language = "English",

volume = "93",

pages = "54--66",

editor = "Olgierd Unold and Witold Dyrka and Wojciech Wieczorek",

booktitle = "Proceedings of Machine Learning Research",

publisher = "PMLR",

url = "http://icgi2018.pwr.edu.pl/papers",

}

TY - GEN

T1 - Learning Product Automata

AU - Moerman, Joshua

N1 - Conference code: 14

PY - 2019

Y1 - 2019

N2 - We give an optimisation for active learning algorithms, applicable to learning Moore machines with decomposable outputs. These machines can be decomposed themselves by projecting on each output. This results in smaller components that can then be learnt with fewer queries. We give experimental evidence that this is a useful technique which can reduce the number of queries substantially. Only in some cases the performance is worsened by the slight overhead. Compositional methods are widely used throughout engineering, and the decomposition presented in this article promises to be particularly interesting for learning hardware systems.

AB - We give an optimisation for active learning algorithms, applicable to learning Moore machines with decomposable outputs. These machines can be decomposed themselves by projecting on each output. This results in smaller components that can then be learnt with fewer queries. We give experimental evidence that this is a useful technique which can reduce the number of queries substantially. Only in some cases the performance is worsened by the slight overhead. Compositional methods are widely used throughout engineering, and the decomposition presented in this article promises to be particularly interesting for learning hardware systems.

M3 - Conference Article in proceeding

VL - 93

SP - 54

EP - 66

BT - Proceedings of Machine Learning Research

A2 - Unold, Olgierd

A2 - Dyrka, Witold

A2 - Wieczorek, Wojciech

PB - PMLR

T2 - The 14th International Conference on Grammatical Inference

Y2 - 5 September 2018 through 7 September 2018

ER -

Learning Product Automata

Abstract

Conference

Access to Document

Fingerprint

Research output

Learning Product Automata

Cite this