Matrix rank minimization approach to signal recovery and nonlinear function estimation for nonlinear ARX model with input nonlinearity

Katsumi Konishi, Masashi Fujii, Katsuyuki Kunida, Shinsuke Uda, Shinya Kuroda

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper deals with an input/output signal recovery problem for nonlinear multiple-input single-output autoregressive exogenous (ARX) models with input nonlinearity, which are used in data-driven systems biology. A matrix rank minimization approach is applied, and a new signal recovery algorithm for nonlinear ARX models is provided. The proposed algorithm recovers output signals and nonlinear-transformed input signals on a linear subspace using some assumptions about nonlinear functions and does not require the exact knowledge of nonlinear functions. Numerical examples using experimental data of signal transduction of cellular systems show the efficiency of the proposed algorithm.
Original languageEnglish
Title of host publicationControl Conference (ASCC), 2017 11th Asian
PublisherIEEE
DOIs
Publication statusPublished - Feb 8 2018

Fingerprint

Recovery
Signal transduction
Systems Biology

Cite this

Matrix rank minimization approach to signal recovery and nonlinear function estimation for nonlinear ARX model with input nonlinearity. / Konishi, Katsumi; Fujii, Masashi; Kunida, Katsuyuki; Uda, Shinsuke; Kuroda, Shinya.

Control Conference (ASCC), 2017 11th Asian. IEEE, 2018.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Konishi, Katsumi ; Fujii, Masashi ; Kunida, Katsuyuki ; Uda, Shinsuke ; Kuroda, Shinya. / Matrix rank minimization approach to signal recovery and nonlinear function estimation for nonlinear ARX model with input nonlinearity. Control Conference (ASCC), 2017 11th Asian. IEEE, 2018.
@inproceedings{1e7f2fd570ed4d63b4dfaa6972c26ada,
title = "Matrix rank minimization approach to signal recovery and nonlinear function estimation for nonlinear ARX model with input nonlinearity",
abstract = "This paper deals with an input/output signal recovery problem for nonlinear multiple-input single-output autoregressive exogenous (ARX) models with input nonlinearity, which are used in data-driven systems biology. A matrix rank minimization approach is applied, and a new signal recovery algorithm for nonlinear ARX models is provided. The proposed algorithm recovers output signals and nonlinear-transformed input signals on a linear subspace using some assumptions about nonlinear functions and does not require the exact knowledge of nonlinear functions. Numerical examples using experimental data of signal transduction of cellular systems show the efficiency of the proposed algorithm.",
author = "Katsumi Konishi and Masashi Fujii and Katsuyuki Kunida and Shinsuke Uda and Shinya Kuroda",
year = "2018",
month = "2",
day = "8",
doi = "10.1109/ASCC.2017.8287382",
language = "English",
booktitle = "Control Conference (ASCC), 2017 11th Asian",
publisher = "IEEE",

}

TY - GEN

T1 - Matrix rank minimization approach to signal recovery and nonlinear function estimation for nonlinear ARX model with input nonlinearity

AU - Konishi, Katsumi

AU - Fujii, Masashi

AU - Kunida, Katsuyuki

AU - Uda, Shinsuke

AU - Kuroda, Shinya

PY - 2018/2/8

Y1 - 2018/2/8

N2 - This paper deals with an input/output signal recovery problem for nonlinear multiple-input single-output autoregressive exogenous (ARX) models with input nonlinearity, which are used in data-driven systems biology. A matrix rank minimization approach is applied, and a new signal recovery algorithm for nonlinear ARX models is provided. The proposed algorithm recovers output signals and nonlinear-transformed input signals on a linear subspace using some assumptions about nonlinear functions and does not require the exact knowledge of nonlinear functions. Numerical examples using experimental data of signal transduction of cellular systems show the efficiency of the proposed algorithm.

AB - This paper deals with an input/output signal recovery problem for nonlinear multiple-input single-output autoregressive exogenous (ARX) models with input nonlinearity, which are used in data-driven systems biology. A matrix rank minimization approach is applied, and a new signal recovery algorithm for nonlinear ARX models is provided. The proposed algorithm recovers output signals and nonlinear-transformed input signals on a linear subspace using some assumptions about nonlinear functions and does not require the exact knowledge of nonlinear functions. Numerical examples using experimental data of signal transduction of cellular systems show the efficiency of the proposed algorithm.

U2 - 10.1109/ASCC.2017.8287382

DO - 10.1109/ASCC.2017.8287382

M3 - Conference contribution

BT - Control Conference (ASCC), 2017 11th Asian

PB - IEEE

ER -