Functional outlier detection and taxonomy by sequential transformations

Wenlin Dai; Tomáš Mrkvička; Ying Sun; Marc G. Genton

doi:10.1016/j.csda.2020.106960

Functional outlier detection and taxonomy by sequential transformations

Wenlin Dai, Tomáš Mrkvička, Ying Sun, Marc G. Genton

Research output: Contribution to journal › Article › peer-review

31 Scopus citations

Abstract

Functional data analysis can be seriously impaired by abnormal observations, which can be classified as either magnitude or shape outliers based on their way of deviating from the bulk of data. Identifying magnitude outliers is relatively easy, while detecting shape outliers is much more challenging. We propose turning the shape outliers into magnitude outliers through data transformation and detecting them using the functional boxplot. Besides easing the detection procedure, applying several transformations sequentially provides a reasonable taxonomy for the flagged outliers. A joint functional ranking, which consists of several transformations, is also defined here. Simulation studies are carried out to evaluate the performance of the proposed method using different functional depth notions. Interesting results are obtained in several practical applications.

Original language	English (US)
Pages (from-to)	106960
Journal	Computational Statistics and Data Analysis
Volume	149
DOIs	https://doi.org/10.1016/j.csda.2020.106960
State	Published - Apr 3 2020

Bibliographical note

KAUST Repository Item: Exported on 2020-10-01
Acknowledgements: This research was supported by the King Abdullah University of Science and Technology (KAUST). Wenlin Dai is also supported by the National Natural Science Foundation of China (Grant No. 11901573).

Access to Document

10.1016/j.csda.2020.106960

https://repository.kaust.edu.sa/bitstream/10754/661060/1/1808.05414.pdf

Cite this

@article{8277df06be74492a9f5248b70174b45e,

title = "Functional outlier detection and taxonomy by sequential transformations",

abstract = "Functional data analysis can be seriously impaired by abnormal observations, which can be classified as either magnitude or shape outliers based on their way of deviating from the bulk of data. Identifying magnitude outliers is relatively easy, while detecting shape outliers is much more challenging. We propose turning the shape outliers into magnitude outliers through data transformation and detecting them using the functional boxplot. Besides easing the detection procedure, applying several transformations sequentially provides a reasonable taxonomy for the flagged outliers. A joint functional ranking, which consists of several transformations, is also defined here. Simulation studies are carried out to evaluate the performance of the proposed method using different functional depth notions. Interesting results are obtained in several practical applications.",

author = "Wenlin Dai and Tom{\'a}{\v s} Mrkvi{\v c}ka and Ying Sun and Genton, {Marc G.}",

note = "KAUST Repository Item: Exported on 2020-10-01 Acknowledgements: This research was supported by the King Abdullah University of Science and Technology (KAUST). Wenlin Dai is also supported by the National Natural Science Foundation of China (Grant No. 11901573).",

year = "2020",

month = apr,

day = "3",

doi = "10.1016/j.csda.2020.106960",

language = "English (US)",

volume = "149",

pages = "106960",

journal = "Computational Statistics and Data Analysis",

issn = "0167-9473",

publisher = "Elsevier",

}

TY - JOUR

T1 - Functional outlier detection and taxonomy by sequential transformations

AU - Dai, Wenlin

AU - Mrkvička, Tomáš

AU - Sun, Ying

AU - Genton, Marc G.

N1 - KAUST Repository Item: Exported on 2020-10-01 Acknowledgements: This research was supported by the King Abdullah University of Science and Technology (KAUST). Wenlin Dai is also supported by the National Natural Science Foundation of China (Grant No. 11901573).

PY - 2020/4/3

Y1 - 2020/4/3

N2 - Functional data analysis can be seriously impaired by abnormal observations, which can be classified as either magnitude or shape outliers based on their way of deviating from the bulk of data. Identifying magnitude outliers is relatively easy, while detecting shape outliers is much more challenging. We propose turning the shape outliers into magnitude outliers through data transformation and detecting them using the functional boxplot. Besides easing the detection procedure, applying several transformations sequentially provides a reasonable taxonomy for the flagged outliers. A joint functional ranking, which consists of several transformations, is also defined here. Simulation studies are carried out to evaluate the performance of the proposed method using different functional depth notions. Interesting results are obtained in several practical applications.

AB - Functional data analysis can be seriously impaired by abnormal observations, which can be classified as either magnitude or shape outliers based on their way of deviating from the bulk of data. Identifying magnitude outliers is relatively easy, while detecting shape outliers is much more challenging. We propose turning the shape outliers into magnitude outliers through data transformation and detecting them using the functional boxplot. Besides easing the detection procedure, applying several transformations sequentially provides a reasonable taxonomy for the flagged outliers. A joint functional ranking, which consists of several transformations, is also defined here. Simulation studies are carried out to evaluate the performance of the proposed method using different functional depth notions. Interesting results are obtained in several practical applications.

UR - http://hdl.handle.net/10754/661060

UR - https://linkinghub.elsevier.com/retrieve/pii/S0167947320300517

UR - http://www.scopus.com/inward/record.url?scp=85083312381&partnerID=8YFLogxK

U2 - 10.1016/j.csda.2020.106960

DO - 10.1016/j.csda.2020.106960

M3 - Article

SN - 0167-9473

VL - 149

SP - 106960

JO - Computational Statistics and Data Analysis

JF - Computational Statistics and Data Analysis

ER -

Functional outlier detection and taxonomy by sequential transformations

Abstract

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this