Feature engineering based on sensory data

Mark Hoogendoorn, Burkhardt Funk

Research output: Chapter in Book / Report / Conference proceedingChapterAcademicpeer-review

Abstract

Approaches to automatically generate useful features from sensory data are introduced in this chapter. Most of the approaches introduced focus on datasets that have a temporal ordering. Features in the time domain are explained, thereby summarizing both numerical and categorical values in a certain historical window. The frequency domain is also discussed, including Fourier transformations and features one can derive from these transformations. In addition, the extraction of features from unstructured data is discussed, mainly focusing on text data.

Original languageEnglish
Title of host publicationMachine Learning for the Quantified Self
Subtitle of host publicationOn the Art of Learning from Sensory Data
PublisherSpringer/Verlag
Chapter4
Pages51-70
Number of pages20
ISBN (Electronic)9783319663081
ISBN (Print)9783319663074
DOIs
Publication statusE-pub ahead of print - 27 Sep 2017

Publication series

NameCognitive Systems Monographs
Volume35
ISSN (Print)1867-4925
ISSN (Electronic)1867-4933

Cite this

Hoogendoorn, M., & Funk, B. (2017). Feature engineering based on sensory data. In Machine Learning for the Quantified Self : On the Art of Learning from Sensory Data (pp. 51-70). (Cognitive Systems Monographs; Vol. 35). Springer/Verlag. https://doi.org/10.1007/978-3-319-66308-1_4
Hoogendoorn, Mark ; Funk, Burkhardt. / Feature engineering based on sensory data. Machine Learning for the Quantified Self : On the Art of Learning from Sensory Data. Springer/Verlag, 2017. pp. 51-70 (Cognitive Systems Monographs).
@inbook{0c68eead210e4fb8863b41eb8d397918,
title = "Feature engineering based on sensory data",
abstract = "Approaches to automatically generate useful features from sensory data are introduced in this chapter. Most of the approaches introduced focus on datasets that have a temporal ordering. Features in the time domain are explained, thereby summarizing both numerical and categorical values in a certain historical window. The frequency domain is also discussed, including Fourier transformations and features one can derive from these transformations. In addition, the extraction of features from unstructured data is discussed, mainly focusing on text data.",
author = "Mark Hoogendoorn and Burkhardt Funk",
year = "2017",
month = "9",
day = "27",
doi = "10.1007/978-3-319-66308-1_4",
language = "English",
isbn = "9783319663074",
series = "Cognitive Systems Monographs",
publisher = "Springer/Verlag",
pages = "51--70",
booktitle = "Machine Learning for the Quantified Self",

}

Hoogendoorn, M & Funk, B 2017, Feature engineering based on sensory data. in Machine Learning for the Quantified Self : On the Art of Learning from Sensory Data. Cognitive Systems Monographs, vol. 35, Springer/Verlag, pp. 51-70. https://doi.org/10.1007/978-3-319-66308-1_4

Feature engineering based on sensory data. / Hoogendoorn, Mark; Funk, Burkhardt.

Machine Learning for the Quantified Self : On the Art of Learning from Sensory Data. Springer/Verlag, 2017. p. 51-70 (Cognitive Systems Monographs; Vol. 35).

Research output: Chapter in Book / Report / Conference proceedingChapterAcademicpeer-review

TY - CHAP

T1 - Feature engineering based on sensory data

AU - Hoogendoorn, Mark

AU - Funk, Burkhardt

PY - 2017/9/27

Y1 - 2017/9/27

N2 - Approaches to automatically generate useful features from sensory data are introduced in this chapter. Most of the approaches introduced focus on datasets that have a temporal ordering. Features in the time domain are explained, thereby summarizing both numerical and categorical values in a certain historical window. The frequency domain is also discussed, including Fourier transformations and features one can derive from these transformations. In addition, the extraction of features from unstructured data is discussed, mainly focusing on text data.

AB - Approaches to automatically generate useful features from sensory data are introduced in this chapter. Most of the approaches introduced focus on datasets that have a temporal ordering. Features in the time domain are explained, thereby summarizing both numerical and categorical values in a certain historical window. The frequency domain is also discussed, including Fourier transformations and features one can derive from these transformations. In addition, the extraction of features from unstructured data is discussed, mainly focusing on text data.

UR - http://www.scopus.com/inward/record.url?scp=85030708142&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85030708142&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-66308-1_4

DO - 10.1007/978-3-319-66308-1_4

M3 - Chapter

SN - 9783319663074

T3 - Cognitive Systems Monographs

SP - 51

EP - 70

BT - Machine Learning for the Quantified Self

PB - Springer/Verlag

ER -

Hoogendoorn M, Funk B. Feature engineering based on sensory data. In Machine Learning for the Quantified Self : On the Art of Learning from Sensory Data. Springer/Verlag. 2017. p. 51-70. (Cognitive Systems Monographs). https://doi.org/10.1007/978-3-319-66308-1_4