Spline-based nonparametric inference in general state-switching models

Roland Langrock; Timo Adam; Vianey Leos-Barajas; Sina Mews; David L. Miller; Yannis P. Papastamatiou

doi:10.1111/stan.12133

Spline-based nonparametric inference in general state-switching models

Roland Langrock^*, Timo Adam, Vianey Leos-Barajas, Sina Mews, David L. Miller, Yannis P. Papastamatiou

^*Corresponding author for this work

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

Abstract

State‐switching models combine immense flexibility with relative mathematical simplicity and computational tractability and, as a consequence, have established themselves as general‐purpose models for time series data. In this paper, we provide an overview of ways to use penalized splines to allow for flexible nonparametric inference within state‐switching models, and provide a critical discussion of the use of corresponding classes of models. The methods are illustrated using animal acceleration data and energy price data.

Original language	English
Pages (from-to)	179-200
Number of pages	22
Journal	Statistica Neerlandica
Volume	72
Issue number	3
Early online date	22 Apr 2018
DOIs	https://doi.org/10.1111/stan.12133
Publication status	Published - 1 Aug 2018
Event	32nd International Workshop on Statistical Modelling (IWSM) - Groningen, Netherlands Duration: 1 Jul 2017 → …

Keywords

Hidden Markov model
Maximum penalized likelihood
Markov-switching regression
Penalized splines

Access to Document

10.1111/stan.12133

statneer-hmm
© 2018, the Authors, Statistica Neerlandica. This work has been made available online in accordance with the publisher’s policies. This is the author created, accepted version manuscript following peer review and may differ slightly from the final published version. The final published version of this work is available at https://doi.org/10.1111/stan.12133
Accepted author manuscript, 1.2 MB

Cite this

@article{be5512a9d9304d45a12ff61650186545,

title = "Spline-based nonparametric inference in general state-switching models",

abstract = "State‐switching models combine immense flexibility with relative mathematical simplicity and computational tractability and, as a consequence, have established themselves as general‐purpose models for time series data. In this paper, we provide an overview of ways to use penalized splines to allow for flexible nonparametric inference within state‐switching models, and provide a critical discussion of the use of corresponding classes of models. The methods are illustrated using animal acceleration data and energy price data. ",

keywords = "Hidden Markov model, Maximum penalized likelihood, Markov-switching regression, Penalized splines",

author = "Roland Langrock and Timo Adam and Vianey Leos-Barajas and Sina Mews and Miller, {David L.} and Papastamatiou, {Yannis P.}",

year = "2018",

month = aug,

day = "1",

doi = "10.1111/stan.12133",

language = "English",

volume = "72",

pages = "179--200",

journal = "Statistica Neerlandica",

issn = "0039-0402",

publisher = "Wiley",

number = "3",

note = "32nd International Workshop on Statistical Modelling (IWSM) ; Conference date: 01-07-2017",

}

TY - JOUR

T1 - Spline-based nonparametric inference in general state-switching models

AU - Langrock, Roland

AU - Adam, Timo

AU - Leos-Barajas, Vianey

AU - Mews, Sina

AU - Miller, David L.

AU - Papastamatiou, Yannis P.

PY - 2018/8/1

Y1 - 2018/8/1

N2 - State‐switching models combine immense flexibility with relative mathematical simplicity and computational tractability and, as a consequence, have established themselves as general‐purpose models for time series data. In this paper, we provide an overview of ways to use penalized splines to allow for flexible nonparametric inference within state‐switching models, and provide a critical discussion of the use of corresponding classes of models. The methods are illustrated using animal acceleration data and energy price data.

AB - State‐switching models combine immense flexibility with relative mathematical simplicity and computational tractability and, as a consequence, have established themselves as general‐purpose models for time series data. In this paper, we provide an overview of ways to use penalized splines to allow for flexible nonparametric inference within state‐switching models, and provide a critical discussion of the use of corresponding classes of models. The methods are illustrated using animal acceleration data and energy price data.

KW - Hidden Markov model

KW - Maximum penalized likelihood

KW - Markov-switching regression

KW - Penalized splines

UR - http://www.scopus.com/inward/record.url?scp=85050095311&partnerID=8YFLogxK

U2 - 10.1111/stan.12133

DO - 10.1111/stan.12133

M3 - Article

SN - 0039-0402

VL - 72

SP - 179

EP - 200

JO - Statistica Neerlandica

JF - Statistica Neerlandica

IS - 3

T2 - 32nd International Workshop on Statistical Modelling (IWSM)

Y2 - 1 July 2017

ER -

Spline-based nonparametric inference in general state-switching models

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this