Combining supervised & unsupervised learning to discover emotional classes

Miguel Arevalillo-Herráez; Aladdin Ayesh; Olga C. Santos; Pablo Arnau-Gonzáalez

doi:10.1145/3079628.3079630

Combining supervised & unsupervised learning to discover emotional classes

Miguel Arevalillo-Herráez, Aladdin Ayesh, Olga C. Santos, Pablo Arnau-Gonzáalez

Research output: Chapter in Book/Report/Conference proceeding › Published conference contribution

1 Citation (Scopus)

Abstract

Most previous work in emotion recognition has fixed the available classes in advance, and a.empted to classify samples into one of these classes using a supervised learning approach. In this paper, we present preliminary work on combining supervised and unsupervised learning to discover potential latent classes which were not initially considered. To illustrate the potential of this hybrid approach, we have used a Self-Organizing Map (SOM) to organize a large number of Electroencephalogram (EEG) signals from subjects watching videos, according to their internal structure. Results suggest that a more useful labelling scheme could be produced by analysing the resulting topology in relation to user reported valence levels (i.e., pleasan tness) for each signal, refining the original set of target classes.

Original language	English
Title of host publication	UMAP 2017 - Proceedings of the 25th Conference on User Modeling, Adaptation and Personalization
Publisher	Association for Computing Machinery, Inc
Pages	355-356
Number of pages	2
ISBN (Electronic)	9781450346351
DOIs	https://doi.org/10.1145/3079628.3079630
Publication status	Published - 9 Jul 2017
Externally published	Yes
Event	25th ACM International Conference on User Modeling, Adaptation, and Personalization, UMAP 2017 - Bratislava, Slovakia Duration: 9 Jul 2017 → 12 Jul 2017

Conference

Conference	25th ACM International Conference on User Modeling, Adaptation, and Personalization, UMAP 2017
Country/Territory	Slovakia
City	Bratislava
Period	9/07/17 → 12/07/17

Bibliographical note

Publisher Copyright:
©2017 ACM.

Keywords

Affective computing
Class discovery
Cluster analysis
EEG
Personalization
User modelling

Access to Document

10.1145/3079628.3079630

Cite this

Combining supervised & unsupervised learning to discover emotional classes. / Arevalillo-Herráez, Miguel; Ayesh, Aladdin; Santos, Olga C. et al.
UMAP 2017 - Proceedings of the 25th Conference on User Modeling, Adaptation and Personalization. Association for Computing Machinery, Inc, 2017. p. 355-356.

Research output: Chapter in Book/Report/Conference proceeding › Published conference contribution

Arevalillo-Herráez, M, Ayesh, A, Santos, OC & Arnau-Gonzáalez, P 2017, Combining supervised & unsupervised learning to discover emotional classes. in UMAP 2017 - Proceedings of the 25th Conference on User Modeling, Adaptation and Personalization. Association for Computing Machinery, Inc, pp. 355-356, 25th ACM International Conference on User Modeling, Adaptation, and Personalization, UMAP 2017, Bratislava, Slovakia, 9/07/17. https://doi.org/10.1145/3079628.3079630

@inproceedings{477351a215dc4317a6182dadd8dd4128,

title = "Combining supervised & unsupervised learning to discover emotional classes",

abstract = "Most previous work in emotion recognition has fixed the available classes in advance, and a.empted to classify samples into one of these classes using a supervised learning approach. In this paper, we present preliminary work on combining supervised and unsupervised learning to discover potential latent classes which were not initially considered. To illustrate the potential of this hybrid approach, we have used a Self-Organizing Map (SOM) to organize a large number of Electroencephalogram (EEG) signals from subjects watching videos, according to their internal structure. Results suggest that a more useful labelling scheme could be produced by analysing the resulting topology in relation to user reported valence levels (i.e., pleasan tness) for each signal, refining the original set of target classes.",

keywords = "Affective computing, Class discovery, Cluster analysis, EEG, Personalization, User modelling",

author = "Miguel Arevalillo-Herr{\'a}ez and Aladdin Ayesh and Santos, {Olga C.} and Pablo Arnau-Gonz{\'a}alez",

note = "Publisher Copyright: {\textcopyright}2017 ACM.; 25th ACM International Conference on User Modeling, Adaptation, and Personalization, UMAP 2017 ; Conference date: 09-07-2017 Through 12-07-2017",

year = "2017",

month = jul,

day = "9",

doi = "10.1145/3079628.3079630",

language = "English",

pages = "355--356",

booktitle = "UMAP 2017 - Proceedings of the 25th Conference on User Modeling, Adaptation and Personalization",

publisher = "Association for Computing Machinery, Inc",

}

TY - GEN

T1 - Combining supervised & unsupervised learning to discover emotional classes

AU - Arevalillo-Herráez, Miguel

AU - Ayesh, Aladdin

AU - Santos, Olga C.

AU - Arnau-Gonzáalez, Pablo

PY - 2017/7/9

Y1 - 2017/7/9

N2 - Most previous work in emotion recognition has fixed the available classes in advance, and a.empted to classify samples into one of these classes using a supervised learning approach. In this paper, we present preliminary work on combining supervised and unsupervised learning to discover potential latent classes which were not initially considered. To illustrate the potential of this hybrid approach, we have used a Self-Organizing Map (SOM) to organize a large number of Electroencephalogram (EEG) signals from subjects watching videos, according to their internal structure. Results suggest that a more useful labelling scheme could be produced by analysing the resulting topology in relation to user reported valence levels (i.e., pleasan tness) for each signal, refining the original set of target classes.

AB - Most previous work in emotion recognition has fixed the available classes in advance, and a.empted to classify samples into one of these classes using a supervised learning approach. In this paper, we present preliminary work on combining supervised and unsupervised learning to discover potential latent classes which were not initially considered. To illustrate the potential of this hybrid approach, we have used a Self-Organizing Map (SOM) to organize a large number of Electroencephalogram (EEG) signals from subjects watching videos, according to their internal structure. Results suggest that a more useful labelling scheme could be produced by analysing the resulting topology in relation to user reported valence levels (i.e., pleasan tness) for each signal, refining the original set of target classes.

KW - Affective computing

KW - Class discovery

KW - Cluster analysis

KW - EEG

KW - Personalization

KW - User modelling

UR - http://www.scopus.com/inward/record.url?scp=85026876705&partnerID=8YFLogxK

U2 - 10.1145/3079628.3079630

DO - 10.1145/3079628.3079630

M3 - Published conference contribution

AN - SCOPUS:85026876705

SP - 355

EP - 356

BT - UMAP 2017 - Proceedings of the 25th Conference on User Modeling, Adaptation and Personalization

PB - Association for Computing Machinery, Inc

T2 - 25th ACM International Conference on User Modeling, Adaptation, and Personalization, UMAP 2017

Y2 - 9 July 2017 through 12 July 2017

ER -

Combining supervised & unsupervised learning to discover emotional classes

Abstract

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this