Deep metric learning for sequential data using approximate information

Stefan Thaler; Vlado Menkovski; Milan Petkovic

doi:10.1007/978-3-319-96136-1_22

Deep metric learning for sequential data using approximate information

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

1 Citation (Scopus)

2 Downloads (Pure)

Abstract

Learning a distance metric provides solutions to many problems where the data exists in a high dimensional space and hand-crafted distance metrics fail to capture its semantical structure. Methods based on deep neural networks such as Siamese or Triplet networks have been developed for learning such metrics. In this paper we present a metric learning method for sequence data based on a RNN-based triplet network. We posit that this model can be trained efficiently with regards to labels by using Jaccard distance as a proxy distance metric. We empirically demonstrate the performance and efficiency of the approach on three different computer log-line datasets.

Original language	English
Title of host publication	Machine Learning and Data Mining in Pattern Recognition - 14th International Conference, MLDM 2018, Proceedings
Editors	Petra Perner
Place of Publication	Cham
Publisher	Springer
Pages	269-282
Number of pages	14
ISBN (Electronic)	978-3-319-96136-1
ISBN (Print)	978-3-319-96135-4
DOIs	https://doi.org/10.1007/978-3-319-96136-1_22
Publication status	Published - 8 Jul 2018
Event	14th International Conference on Machine Learning and Data Mining in Pattern Recognition, MLDM 2018 - New York, United States Duration: 15 Jul 2018 → 19 Jul 2018

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10934 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	14th International Conference on Machine Learning and Data Mining in Pattern Recognition, MLDM 2018
Country/Territory	United States
City	New York
Period	15/07/18 → 19/07/18

Funding

Acknowledgment. The work presented in this paper is part of a project which has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No. 780495.

Keywords

Deep learning
Efficient metric learning
Triplet network

Access to Document

10.1007/978-3-319-96136-1_22

Cite this

Thaler, S., Menkovski, V., & Petkovic, M. (2018). Deep metric learning for sequential data using approximate information. In P. Perner (Ed.), Machine Learning and Data Mining in Pattern Recognition - 14th International Conference, MLDM 2018, Proceedings (pp. 269-282). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10934 LNAI). Springer. https://doi.org/10.1007/978-3-319-96136-1_22

Thaler, Stefan ; Menkovski, Vlado ; Petkovic, Milan. / Deep metric learning for sequential data using approximate information. Machine Learning and Data Mining in Pattern Recognition - 14th International Conference, MLDM 2018, Proceedings. editor / Petra Perner. Cham : Springer, 2018. pp. 269-282 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{1a812ec0bf7f429f9ccd5650a92bff64,

title = "Deep metric learning for sequential data using approximate information",

abstract = "Learning a distance metric provides solutions to many problems where the data exists in a high dimensional space and hand-crafted distance metrics fail to capture its semantical structure. Methods based on deep neural networks such as Siamese or Triplet networks have been developed for learning such metrics. In this paper we present a metric learning method for sequence data based on a RNN-based triplet network. We posit that this model can be trained efficiently with regards to labels by using Jaccard distance as a proxy distance metric. We empirically demonstrate the performance and efficiency of the approach on three different computer log-line datasets.",

keywords = "Deep learning, Efficient metric learning, Triplet network",

author = "Stefan Thaler and Vlado Menkovski and Milan Petkovic",

year = "2018",

month = jul,

day = "8",

doi = "10.1007/978-3-319-96136-1\_22",

language = "English",

isbn = "978-3-319-96135-4",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer",

pages = "269--282",

editor = "Petra Perner",

booktitle = "Machine Learning and Data Mining in Pattern Recognition - 14th International Conference, MLDM 2018, Proceedings",

address = "Germany",

note = "14th International Conference on Machine Learning and Data Mining in Pattern Recognition, MLDM 2018 ; Conference date: 15-07-2018 Through 19-07-2018",

}

Thaler, S, Menkovski, V & Petkovic, M 2018, Deep metric learning for sequential data using approximate information. in P Perner (ed.), Machine Learning and Data Mining in Pattern Recognition - 14th International Conference, MLDM 2018, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10934 LNAI, Springer, Cham, pp. 269-282, 14th International Conference on Machine Learning and Data Mining in Pattern Recognition, MLDM 2018, New York, United States, 15/07/18. https://doi.org/10.1007/978-3-319-96136-1_22

Deep metric learning for sequential data using approximate information. / Thaler, Stefan; Menkovski, Vlado ; Petkovic, Milan.
Machine Learning and Data Mining in Pattern Recognition - 14th International Conference, MLDM 2018, Proceedings. ed. / Petra Perner. Cham: Springer, 2018. p. 269-282 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10934 LNAI).

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

TY - GEN

T1 - Deep metric learning for sequential data using approximate information

AU - Thaler, Stefan

AU - Menkovski, Vlado

AU - Petkovic, Milan

PY - 2018/7/8

Y1 - 2018/7/8

N2 - Learning a distance metric provides solutions to many problems where the data exists in a high dimensional space and hand-crafted distance metrics fail to capture its semantical structure. Methods based on deep neural networks such as Siamese or Triplet networks have been developed for learning such metrics. In this paper we present a metric learning method for sequence data based on a RNN-based triplet network. We posit that this model can be trained efficiently with regards to labels by using Jaccard distance as a proxy distance metric. We empirically demonstrate the performance and efficiency of the approach on three different computer log-line datasets.

AB - Learning a distance metric provides solutions to many problems where the data exists in a high dimensional space and hand-crafted distance metrics fail to capture its semantical structure. Methods based on deep neural networks such as Siamese or Triplet networks have been developed for learning such metrics. In this paper we present a metric learning method for sequence data based on a RNN-based triplet network. We posit that this model can be trained efficiently with regards to labels by using Jaccard distance as a proxy distance metric. We empirically demonstrate the performance and efficiency of the approach on three different computer log-line datasets.

KW - Deep learning

KW - Efficient metric learning

KW - Triplet network

UR - https://www.scopus.com/pages/publications/85050538775

U2 - 10.1007/978-3-319-96136-1_22

DO - 10.1007/978-3-319-96136-1_22

M3 - Conference contribution

AN - SCOPUS:85050538775

SN - 978-3-319-96135-4

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 269

EP - 282

BT - Machine Learning and Data Mining in Pattern Recognition - 14th International Conference, MLDM 2018, Proceedings

A2 - Perner, Petra

PB - Springer

CY - Cham

T2 - 14th International Conference on Machine Learning and Data Mining in Pattern Recognition, MLDM 2018

Y2 - 15 July 2018 through 19 July 2018

ER -

Thaler S, Menkovski V , Petkovic M. Deep metric learning for sequential data using approximate information. In Perner P, editor, Machine Learning and Data Mining in Pattern Recognition - 14th International Conference, MLDM 2018, Proceedings. Cham: Springer. 2018. p. 269-282. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-96136-1_22

Deep metric learning for sequential data using approximate information

Abstract

Publication series

Conference

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this