Support vector mind map of wine speak

Brendan Flanagan, Sachio Hirokawa

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Models created by blackbox machine learning techniques such as SVM can be difficult to interpret. It is because these methods do not offer a clear explanation of how classifications are derived that is easy for humans to understand. Other machine learning techniques, such as: decision trees, produce models that are intuitive for humans to interpret. However, there are often cases where an SVM model will out preform a more intuitive model, making interpretation of SVM trained models an important problem. In this paper, we propose a method of visualizing linear SVM models for text classification by analyzing the relation of features in the support vectors. An example of this method is shown in a case study into the interpretation of a model trained on wine tasting notes.

Original languageEnglish
Title of host publicationHuman Interface and the Management of Information
Subtitle of host publicationInformation, Design and Interaction - 18th International Conference, HCI International 2016, Proceedings
EditorsSakae Yamamoto
PublisherSpringer Verlag
Pages127-135
Number of pages9
ISBN (Print)9783319403489
DOIs
Publication statusPublished - Jan 1 2016
Event18th International Conference on Human-Computer Interaction, HCI International 2016 - Toronto, Canada
Duration: Jul 17 2016Jul 22 2016

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume9734
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other18th International Conference on Human-Computer Interaction, HCI International 2016
CountryCanada
CityToronto
Period7/17/167/22/16

Fingerprint

Wine
Support Vector
Learning systems
Intuitive
Model
Machine Learning
Text Classification
Decision trees
Black Box
Decision tree

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Flanagan, B., & Hirokawa, S. (2016). Support vector mind map of wine speak. In S. Yamamoto (Ed.), Human Interface and the Management of Information: Information, Design and Interaction - 18th International Conference, HCI International 2016, Proceedings (pp. 127-135). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9734). Springer Verlag. https://doi.org/10.1007/978-3-319-40349-6_13

Support vector mind map of wine speak. / Flanagan, Brendan; Hirokawa, Sachio.

Human Interface and the Management of Information: Information, Design and Interaction - 18th International Conference, HCI International 2016, Proceedings. ed. / Sakae Yamamoto. Springer Verlag, 2016. p. 127-135 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9734).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Flanagan, B & Hirokawa, S 2016, Support vector mind map of wine speak. in S Yamamoto (ed.), Human Interface and the Management of Information: Information, Design and Interaction - 18th International Conference, HCI International 2016, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9734, Springer Verlag, pp. 127-135, 18th International Conference on Human-Computer Interaction, HCI International 2016, Toronto, Canada, 7/17/16. https://doi.org/10.1007/978-3-319-40349-6_13
Flanagan B, Hirokawa S. Support vector mind map of wine speak. In Yamamoto S, editor, Human Interface and the Management of Information: Information, Design and Interaction - 18th International Conference, HCI International 2016, Proceedings. Springer Verlag. 2016. p. 127-135. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-40349-6_13
Flanagan, Brendan ; Hirokawa, Sachio. / Support vector mind map of wine speak. Human Interface and the Management of Information: Information, Design and Interaction - 18th International Conference, HCI International 2016, Proceedings. editor / Sakae Yamamoto. Springer Verlag, 2016. pp. 127-135 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{c434c2a194d140198bfb970c803a5578,
title = "Support vector mind map of wine speak",
abstract = "Models created by blackbox machine learning techniques such as SVM can be difficult to interpret. It is because these methods do not offer a clear explanation of how classifications are derived that is easy for humans to understand. Other machine learning techniques, such as: decision trees, produce models that are intuitive for humans to interpret. However, there are often cases where an SVM model will out preform a more intuitive model, making interpretation of SVM trained models an important problem. In this paper, we propose a method of visualizing linear SVM models for text classification by analyzing the relation of features in the support vectors. An example of this method is shown in a case study into the interpretation of a model trained on wine tasting notes.",
author = "Brendan Flanagan and Sachio Hirokawa",
year = "2016",
month = "1",
day = "1",
doi = "10.1007/978-3-319-40349-6_13",
language = "English",
isbn = "9783319403489",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "127--135",
editor = "Sakae Yamamoto",
booktitle = "Human Interface and the Management of Information",
address = "Germany",

}

TY - GEN

T1 - Support vector mind map of wine speak

AU - Flanagan, Brendan

AU - Hirokawa, Sachio

PY - 2016/1/1

Y1 - 2016/1/1

N2 - Models created by blackbox machine learning techniques such as SVM can be difficult to interpret. It is because these methods do not offer a clear explanation of how classifications are derived that is easy for humans to understand. Other machine learning techniques, such as: decision trees, produce models that are intuitive for humans to interpret. However, there are often cases where an SVM model will out preform a more intuitive model, making interpretation of SVM trained models an important problem. In this paper, we propose a method of visualizing linear SVM models for text classification by analyzing the relation of features in the support vectors. An example of this method is shown in a case study into the interpretation of a model trained on wine tasting notes.

AB - Models created by blackbox machine learning techniques such as SVM can be difficult to interpret. It is because these methods do not offer a clear explanation of how classifications are derived that is easy for humans to understand. Other machine learning techniques, such as: decision trees, produce models that are intuitive for humans to interpret. However, there are often cases where an SVM model will out preform a more intuitive model, making interpretation of SVM trained models an important problem. In this paper, we propose a method of visualizing linear SVM models for text classification by analyzing the relation of features in the support vectors. An example of this method is shown in a case study into the interpretation of a model trained on wine tasting notes.

UR - http://www.scopus.com/inward/record.url?scp=84978785910&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84978785910&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-40349-6_13

DO - 10.1007/978-3-319-40349-6_13

M3 - Conference contribution

AN - SCOPUS:84978785910

SN - 9783319403489

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

SP - 127

EP - 135

BT - Human Interface and the Management of Information

A2 - Yamamoto, Sakae

PB - Springer Verlag

ER -