Piecewise nonlinear regression with data augmentation

M. Mazzoleni; V. Breschi; S. Formentin

doi:10.1016/j.ifacol.2021.08.396

Piecewise nonlinear regression with data augmentation

M. Mazzoleni
, V. Breschi
, S. Formentin

Research output: Contribution to journal › Conference article › peer-review

6 Citations (Scopus)

Abstract

Piecewise regression represents a powerful tool to derive accurate yet modular models describing complex phenomena or physical systems. This paper presents an approach for learning PieceWise NonLinear (PWNL) functions in both a supervised and semi-supervised setting. We further equip the proposed technique with a method for the automatic generation of additional unsupervised data, which are leveraged to improve the overall accuracy of the estimate. The performance of the proposed approach is preliminarily assessed on two simple simulation examples, where we show the benefits of using nonlinear local models and artificially generated unsupervised data.

Original language	English
Pages (from-to)	421-426
Number of pages	6
Journal	IFAC-PapersOnLine
Volume	54
Issue number	7
DOIs	https://doi.org/10.1016/j.ifacol.2021.08.396
Publication status	Published - 1 Jul 2021
Externally published	Yes
Event	19th IFAC Symposium on System Identification (SYSID 2021) - Virtual, Padova, Italy Duration: 13 Jul 2021 → 16 Jul 2021 Conference number: 19 https://www.sysid2021.org/

Keywords

Hybrid System Identification
Nonlinear System Identification
Nonparametric Methods

Access to Document

10.1016/j.ifacol.2021.08.396

Cite this

@article{d1502f564c854482aed38e8db5f72618,

title = "Piecewise nonlinear regression with data augmentation",

abstract = "Piecewise regression represents a powerful tool to derive accurate yet modular models describing complex phenomena or physical systems. This paper presents an approach for learning PieceWise NonLinear (PWNL) functions in both a supervised and semi-supervised setting. We further equip the proposed technique with a method for the automatic generation of additional unsupervised data, which are leveraged to improve the overall accuracy of the estimate. The performance of the proposed approach is preliminarily assessed on two simple simulation examples, where we show the benefits of using nonlinear local models and artificially generated unsupervised data.",

keywords = "Hybrid System Identification, Nonlinear System Identification, Nonparametric Methods",

author = "M. Mazzoleni and V. Breschi and S. Formentin",

note = "Publisher Copyright: Copyright {\textcopyright} 2021 The Authors.; 19th IFAC Symposium on System Identification (SYSID 2021), SYSID 2021 ; Conference date: 13-07-2021 Through 16-07-2021",

year = "2021",

month = jul,

day = "1",

doi = "10.1016/j.ifacol.2021.08.396",

language = "English",

volume = "54",

pages = "421--426",

journal = "IFAC-PapersOnLine",

issn = "2405-8963",

publisher = "Elsevier",

number = "7",

url = "https://www.sysid2021.org/",

}

TY - JOUR

T1 - Piecewise nonlinear regression with data augmentation

AU - Mazzoleni, M.

AU - Breschi, V.

AU - Formentin, S.

N1 - Conference code: 19

PY - 2021/7/1

Y1 - 2021/7/1

N2 - Piecewise regression represents a powerful tool to derive accurate yet modular models describing complex phenomena or physical systems. This paper presents an approach for learning PieceWise NonLinear (PWNL) functions in both a supervised and semi-supervised setting. We further equip the proposed technique with a method for the automatic generation of additional unsupervised data, which are leveraged to improve the overall accuracy of the estimate. The performance of the proposed approach is preliminarily assessed on two simple simulation examples, where we show the benefits of using nonlinear local models and artificially generated unsupervised data.

AB - Piecewise regression represents a powerful tool to derive accurate yet modular models describing complex phenomena or physical systems. This paper presents an approach for learning PieceWise NonLinear (PWNL) functions in both a supervised and semi-supervised setting. We further equip the proposed technique with a method for the automatic generation of additional unsupervised data, which are leveraged to improve the overall accuracy of the estimate. The performance of the proposed approach is preliminarily assessed on two simple simulation examples, where we show the benefits of using nonlinear local models and artificially generated unsupervised data.

KW - Hybrid System Identification

KW - Nonlinear System Identification

KW - Nonparametric Methods

UR - https://www.scopus.com/pages/publications/85118108402

U2 - 10.1016/j.ifacol.2021.08.396

DO - 10.1016/j.ifacol.2021.08.396

M3 - Conference article

AN - SCOPUS:85118108402

SN - 2405-8963

VL - 54

SP - 421

EP - 426

JO - IFAC-PapersOnLine

JF - IFAC-PapersOnLine

IS - 7

T2 - 19th IFAC Symposium on System Identification (SYSID 2021)

Y2 - 13 July 2021 through 16 July 2021

ER -

Piecewise nonlinear regression with data augmentation

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this