Lifted Junction Tree Algorithm

Tanya Braun; Ralf Möller

doi:10.1007/978-3-319-46073-4_3

Lifted Junction Tree Algorithm

Tanya Braun, Ralf Möller

Institut für Informationssysteme

44 Zitate (Scopus)

Abstract

We look at probabilistic first-order formalisms where the domain objects are known. In these formalisms, the standard approach for inference is lifted variable elimination. To benefit from the advantages of the junction tree algorithm for inference in the first-order setting, we transfer the idea of lifting to the junction tree algorithm.

Our lifted junction tree algorithm aims at reducing computations by introducing first-order junction trees that compactly represent symmetries. First experiments show that we speed up the computation time compared to the propositional version. When querying for multiple marginals, the lifted junction tree algorithm performs better than using lifted VE to infer each marginal individually.

Originalsprache	Englisch
Titel	KI 2016: Advances in Artificial Intelligence
Redakteure/-innen	Gerhard Friedrich, Malte Helmert, Franz Wotawa
Seitenumfang	13
Band	9904
Erscheinungsort	Cham
Herausgeber (Verlag)	Springer International Publishing
Erscheinungsdatum	08.09.2016
Seiten	30-42
ISBN (Print)	978-3-319-46072-7
ISBN (elektronisch)	978-3-319-46073-4
DOIs	https://doi.org/10.1007/978-3-319-46073-4_3
Publikationsstatus	Veröffentlicht - 08.09.2016
Veranstaltung	39th German Conference on Artificial Intelligence - Klagenfurt, Österreich Dauer: 26.09.2016 → 30.09.2016 Konferenznummer: 181639

UN SDGs

Dieser Output leistet einen Beitrag zu folgendem(n) Ziel(en) für nachhaltige Entwicklung

SDG 3 – Gesundheit und Wohlergehen
SDG 9 – Industrie, Innovation und Infrastruktur

DFG-Fachsystematik

4.43-01 Theoretische Informatik

Dokumente

10.1007/978-3-319-46073-4_3

Weitere Links

https://www.scopus.com/pages/publications/84988600847

Zitieren

@inproceedings{80528f9625ca424c9b7c210d98c36367,

title = "Lifted Junction Tree Algorithm",

abstract = "We look at probabilistic first-order formalisms where the domain objects are known. In these formalisms, the standard approach for inference is lifted variable elimination. To benefit from the advantages of the junction tree algorithm for inference in the first-order setting, we transfer the idea of lifting to the junction tree algorithm. Our lifted junction tree algorithm aims at reducing computations by introducing first-order junction trees that compactly represent symmetries. First experiments show that we speed up the computation time compared to the propositional version. When querying for multiple marginals, the lifted junction tree algorithm performs better than using lifted VE to infer each marginal individually.",

author = "Tanya Braun and Ralf M{\"o}ller",

year = "2016",

month = sep,

day = "8",

doi = "10.1007/978-3-319-46073-4\_3",

language = "English",

isbn = "978-3-319-46072-7",

volume = "9904",

series = "Lecture Notes in Computer Science",

publisher = "Springer International Publishing",

pages = "30--42",

editor = "Gerhard Friedrich and Malte Helmert and Franz Wotawa",

booktitle = "KI 2016: Advances in Artificial Intelligence",

note = "39th German Conference on Artificial Intelligence, KI 2016 ; Conference date: 26-09-2016 Through 30-09-2016",

}

Braun, T & Möller, R 2016, Lifted Junction Tree Algorithm. in G Friedrich, M Helmert & F Wotawa (Hrsg.), KI 2016: Advances in Artificial Intelligence. Bd. 9904, Lecture Notes in Computer Science, Bd. 9904 , Springer International Publishing, Cham, S. 30-42, 39th German Conference on Artificial Intelligence, Klagenfurt, Österreich, 26.09.16. https://doi.org/10.1007/978-3-319-46073-4_3

Lifted Junction Tree Algorithm. / Braun, Tanya; Möller, Ralf.
KI 2016: Advances in Artificial Intelligence. Hrsg. / Gerhard Friedrich; Malte Helmert; Franz Wotawa. Band 9904 Cham: Springer International Publishing, 2016. S. 30-42 (Lecture Notes in Computer Science; Band 9904 ).

Publikation: Kapitel in Büchern/Berichten/Konferenzbänden › Konferenzbeitrag › Begutachtung

TY - GEN

T1 - Lifted Junction Tree Algorithm

AU - Braun, Tanya

AU - Möller, Ralf

N1 - Conference code: 181639

PY - 2016/9/8

Y1 - 2016/9/8

N2 - We look at probabilistic first-order formalisms where the domain objects are known. In these formalisms, the standard approach for inference is lifted variable elimination. To benefit from the advantages of the junction tree algorithm for inference in the first-order setting, we transfer the idea of lifting to the junction tree algorithm. Our lifted junction tree algorithm aims at reducing computations by introducing first-order junction trees that compactly represent symmetries. First experiments show that we speed up the computation time compared to the propositional version. When querying for multiple marginals, the lifted junction tree algorithm performs better than using lifted VE to infer each marginal individually.

AB - We look at probabilistic first-order formalisms where the domain objects are known. In these formalisms, the standard approach for inference is lifted variable elimination. To benefit from the advantages of the junction tree algorithm for inference in the first-order setting, we transfer the idea of lifting to the junction tree algorithm. Our lifted junction tree algorithm aims at reducing computations by introducing first-order junction trees that compactly represent symmetries. First experiments show that we speed up the computation time compared to the propositional version. When querying for multiple marginals, the lifted junction tree algorithm performs better than using lifted VE to infer each marginal individually.

UR - https://www.scopus.com/pages/publications/84988600847

U2 - 10.1007/978-3-319-46073-4_3

DO - 10.1007/978-3-319-46073-4_3

M3 - Conference contribution

SN - 978-3-319-46072-7

VL - 9904

T3 - Lecture Notes in Computer Science

SP - 30

EP - 42

BT - KI 2016: Advances in Artificial Intelligence

A2 - Friedrich, Gerhard

A2 - Helmert, Malte

A2 - Wotawa, Franz

PB - Springer International Publishing

CY - Cham

T2 - 39th German Conference on Artificial Intelligence

Y2 - 26 September 2016 through 30 September 2016

ER -

Lifted Junction Tree Algorithm

Abstract

UN SDGs

DFG-Fachsystematik

Dokumente

Weitere Links

Fingerprint

Zitieren