Smooth noisy PCA using a 1st order roughness penalty

Jakob Sigurdsson; Magnus O. Ulfarsson

doi:10.1109/MLSP.2010.5589208

Smooth noisy PCA using a 1st order roughness penalty

Faculty of Electrical and Computer Engineering

University of Iceland

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

Abstract

Principal component analysis (PCA) and other multivariate methods have proven to be useful in a variety of engineering and science fields. PCA is commonly used for dimensionality reduction. PCA has also proven to be useful in functional magnetic resonance imaging (fMRI) research where it is used to decompose the fMRI data into components which can be associated with biological processes. In this paper we develop a smooth version of PCA derived from a maximum likelihood framework. A 1^st order roughness penalty term is added to the log-likelihood function which is then maximized for the parameters of interest with an expectation maximization (EM) algorithm. This new method is applied both to simulated data and real fMRI data.

Original language	English
Title of host publication	Proceedings of the 2010 IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2010
Pages	325-330
Number of pages	6
DOIs	https://doi.org/10.1109/MLSP.2010.5589208 https://doi.org/10.1109/mlsp.2010.5589208
Publication status	Published - 2010
Event	2010 IEEE 20th International Workshop on Machine Learning for Signal Processing, MLSP 2010 - Kittila, Finland Duration: 29 Aug 2010 → 1 Sept 2010

Publication series

Name	Proceedings of the 2010 IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2010

Conference

Conference	2010 IEEE 20th International Workshop on Machine Learning for Signal Processing, MLSP 2010
Country/Territory	Finland
City	Kittila
Period	29/08/10 → 1/09/10

Access to Document

Cite this

Sigurdsson, J., & Ulfarsson, M. O. (2010). Smooth noisy PCA using a 1st order roughness penalty. In Proceedings of the 2010 IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2010 (pp. 325-330). Article 5589208 (Proceedings of the 2010 IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2010). https://doi.org/10.1109/MLSP.2010.5589208, https://doi.org/10.1109/mlsp.2010.5589208

@inproceedings{68777025cd3d41f5974c9748698cfe98,

title = "Smooth noisy PCA using a 1st order roughness penalty",

abstract = "Principal component analysis (PCA) and other multivariate methods have proven to be useful in a variety of engineering and science fields. PCA is commonly used for dimensionality reduction. PCA has also proven to be useful in functional magnetic resonance imaging (fMRI) research where it is used to decompose the fMRI data into components which can be associated with biological processes. In this paper we develop a smooth version of PCA derived from a maximum likelihood framework. A 1st order roughness penalty term is added to the log-likelihood function which is then maximized for the parameters of interest with an expectation maximization (EM) algorithm. This new method is applied both to simulated data and real fMRI data.",

author = "Jakob Sigurdsson and Ulfarsson, \{Magnus O.\}",

year = "2010",

doi = "10.1109/MLSP.2010.5589208",

language = "English",

isbn = "9781424478774",

series = "Proceedings of the 2010 IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2010",

pages = "325--330",

booktitle = "Proceedings of the 2010 IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2010",

note = "2010 IEEE 20th International Workshop on Machine Learning for Signal Processing, MLSP 2010 ; Conference date: 29-08-2010 Through 01-09-2010",

}

Sigurdsson, J & Ulfarsson, MO 2010, Smooth noisy PCA using a 1st order roughness penalty. in Proceedings of the 2010 IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2010., 5589208, Proceedings of the 2010 IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2010, pp. 325-330, 2010 IEEE 20th International Workshop on Machine Learning for Signal Processing, MLSP 2010, Kittila, Finland, 29/08/10. https://doi.org/10.1109/MLSP.2010.5589208, https://doi.org/10.1109/mlsp.2010.5589208

Smooth noisy PCA using a 1st order roughness penalty. / Sigurdsson, Jakob ; Ulfarsson, Magnus O.
Proceedings of the 2010 IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2010. 2010. p. 325-330 5589208 (Proceedings of the 2010 IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2010).

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

TY - GEN

T1 - Smooth noisy PCA using a 1st order roughness penalty

AU - Sigurdsson, Jakob

AU - Ulfarsson, Magnus O.

PY - 2010

Y1 - 2010

N2 - Principal component analysis (PCA) and other multivariate methods have proven to be useful in a variety of engineering and science fields. PCA is commonly used for dimensionality reduction. PCA has also proven to be useful in functional magnetic resonance imaging (fMRI) research where it is used to decompose the fMRI data into components which can be associated with biological processes. In this paper we develop a smooth version of PCA derived from a maximum likelihood framework. A 1st order roughness penalty term is added to the log-likelihood function which is then maximized for the parameters of interest with an expectation maximization (EM) algorithm. This new method is applied both to simulated data and real fMRI data.

AB - Principal component analysis (PCA) and other multivariate methods have proven to be useful in a variety of engineering and science fields. PCA is commonly used for dimensionality reduction. PCA has also proven to be useful in functional magnetic resonance imaging (fMRI) research where it is used to decompose the fMRI data into components which can be associated with biological processes. In this paper we develop a smooth version of PCA derived from a maximum likelihood framework. A 1st order roughness penalty term is added to the log-likelihood function which is then maximized for the parameters of interest with an expectation maximization (EM) algorithm. This new method is applied both to simulated data and real fMRI data.

UR - https://www.scopus.com/pages/publications/78449312088

U2 - 10.1109/MLSP.2010.5589208

DO - 10.1109/MLSP.2010.5589208

M3 - Conference contribution

SN - 9781424478774

T3 - Proceedings of the 2010 IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2010

SP - 325

EP - 330

BT - Proceedings of the 2010 IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2010

T2 - 2010 IEEE 20th International Workshop on Machine Learning for Signal Processing, MLSP 2010

Y2 - 29 August 2010 through 1 September 2010

ER -

Sigurdsson J , Ulfarsson MO. Smooth noisy PCA using a 1st order roughness penalty. In Proceedings of the 2010 IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2010. 2010. p. 325-330. 5589208. (Proceedings of the 2010 IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2010). doi: 10.1109/MLSP.2010.5589208, 10.1109/mlsp.2010.5589208

Smooth noisy PCA using a 1<sup>st</sup> order roughness penalty

Abstract

Publication series

Conference

Access to Document

Fingerprint

Cite this