Feature selection and activity recognition from wearable sensors

Susanna Pirttikangas; Kaori Fujinami; Tatsuo Nakajima

doi:10.1007/11890348_39

Feature selection and activity recognition from wearable sensors

Susanna Pirttikangas^*, Kaori Fujinami, Tatsuo Nakajima

^*Corresponding author for this work

School of Fundamental Science and Engineering

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

192 Citations (Scopus)

Abstract

We describe our data collection and results on activity recognition with wearable, coin-sized sensor devices. The devices were attached to four different parts of the body: right thigh and wrist, left wrist and to a necklace on 13 different testees. In this experiment, data was from 17 daily life examples from male and female subjects. Features were calculated from triaxial accelerometer and heart rate data within different sized time windows. The best features were selected with forward-back ward sequential search algorithm. Interestingly, acceleration mean values from the necklace were selected as important features. Two classifiers (multilayer perceptrons and kNN classifiers) were tested for activity recognition, and the best result (90.61 % aggregate recognition rate for 4-fold cross validation) was achieved with a kNN classifier.

Original language	English
Title of host publication	Ubiquitous Computing Systems - Third International Symposium, UCS 2006, Proceedings
Publisher	Springer Verlag
Pages	516-527
Number of pages	12
ISBN (Print)	3540462872, 9783540462873
DOIs	https://doi.org/10.1007/11890348_39
Publication status	Published - 2006
Event	3rd International Symposium on Ubiquitous Computing Systems, UCS 2006 - Seoul, Korea, Republic of Duration: 2006 Oct 11 → 2006 Oct 13

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	4239 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	3rd International Symposium on Ubiquitous Computing Systems, UCS 2006
Country/Territory	Korea, Republic of
City	Seoul
Period	06/10/11 → 06/10/13

ASJC Scopus subject areas

Theoretical Computer Science
Computer Science(all)

Access to Document

10.1007/11890348_39

Cite this

Pirttikangas, S., Fujinami, K., & Nakajima, T. (2006). Feature selection and activity recognition from wearable sensors. In Ubiquitous Computing Systems - Third International Symposium, UCS 2006, Proceedings (pp. 516-527). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4239 LNCS). Springer Verlag. https://doi.org/10.1007/11890348_39

Feature selection and activity recognition from wearable sensors. / Pirttikangas, Susanna; Fujinami, Kaori; Nakajima, Tatsuo.
Ubiquitous Computing Systems - Third International Symposium, UCS 2006, Proceedings. Springer Verlag, 2006. p. 516-527 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4239 LNCS).

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

Pirttikangas, S, Fujinami, K & Nakajima, T 2006, Feature selection and activity recognition from wearable sensors. in Ubiquitous Computing Systems - Third International Symposium, UCS 2006, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 4239 LNCS, Springer Verlag, pp. 516-527, 3rd International Symposium on Ubiquitous Computing Systems, UCS 2006, Seoul, Korea, Republic of, 06/10/11. https://doi.org/10.1007/11890348_39

Pirttikangas S, Fujinami K, Nakajima T. Feature selection and activity recognition from wearable sensors. In Ubiquitous Computing Systems - Third International Symposium, UCS 2006, Proceedings. Springer Verlag. 2006. p. 516-527. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/11890348_39

Pirttikangas, Susanna ; Fujinami, Kaori ; Nakajima, Tatsuo. / Feature selection and activity recognition from wearable sensors. Ubiquitous Computing Systems - Third International Symposium, UCS 2006, Proceedings. Springer Verlag, 2006. pp. 516-527 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{4c064aa37d1e466c96771d41d552252c,

title = "Feature selection and activity recognition from wearable sensors",

abstract = "We describe our data collection and results on activity recognition with wearable, coin-sized sensor devices. The devices were attached to four different parts of the body: right thigh and wrist, left wrist and to a necklace on 13 different testees. In this experiment, data was from 17 daily life examples from male and female subjects. Features were calculated from triaxial accelerometer and heart rate data within different sized time windows. The best features were selected with forward-back ward sequential search algorithm. Interestingly, acceleration mean values from the necklace were selected as important features. Two classifiers (multilayer perceptrons and kNN classifiers) were tested for activity recognition, and the best result (90.61 % aggregate recognition rate for 4-fold cross validation) was achieved with a kNN classifier.",

author = "Susanna Pirttikangas and Kaori Fujinami and Tatsuo Nakajima",

year = "2006",

doi = "10.1007/11890348_39",

language = "English",

isbn = "3540462872",

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

publisher = "Springer Verlag",

pages = "516--527",

booktitle = "Ubiquitous Computing Systems - Third International Symposium, UCS 2006, Proceedings",

}

TY - GEN

T1 - Feature selection and activity recognition from wearable sensors

AU - Pirttikangas, Susanna

AU - Fujinami, Kaori

AU - Nakajima, Tatsuo

PY - 2006

Y1 - 2006

N2 - We describe our data collection and results on activity recognition with wearable, coin-sized sensor devices. The devices were attached to four different parts of the body: right thigh and wrist, left wrist and to a necklace on 13 different testees. In this experiment, data was from 17 daily life examples from male and female subjects. Features were calculated from triaxial accelerometer and heart rate data within different sized time windows. The best features were selected with forward-back ward sequential search algorithm. Interestingly, acceleration mean values from the necklace were selected as important features. Two classifiers (multilayer perceptrons and kNN classifiers) were tested for activity recognition, and the best result (90.61 % aggregate recognition rate for 4-fold cross validation) was achieved with a kNN classifier.

AB - We describe our data collection and results on activity recognition with wearable, coin-sized sensor devices. The devices were attached to four different parts of the body: right thigh and wrist, left wrist and to a necklace on 13 different testees. In this experiment, data was from 17 daily life examples from male and female subjects. Features were calculated from triaxial accelerometer and heart rate data within different sized time windows. The best features were selected with forward-back ward sequential search algorithm. Interestingly, acceleration mean values from the necklace were selected as important features. Two classifiers (multilayer perceptrons and kNN classifiers) were tested for activity recognition, and the best result (90.61 % aggregate recognition rate for 4-fold cross validation) was achieved with a kNN classifier.

UR - http://www.scopus.com/inward/record.url?scp=33750707878&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33750707878&partnerID=8YFLogxK

U2 - 10.1007/11890348_39

DO - 10.1007/11890348_39

M3 - Conference contribution

AN - SCOPUS:33750707878

SN - 3540462872

SN - 9783540462873

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

SP - 516

EP - 527

BT - Ubiquitous Computing Systems - Third International Symposium, UCS 2006, Proceedings

PB - Springer Verlag

T2 - 3rd International Symposium on Ubiquitous Computing Systems, UCS 2006

Y2 - 11 October 2006 through 13 October 2006

ER -

Feature selection and activity recognition from wearable sensors

Abstract

Publication series

Conference

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this