Time-Series Analysis Using Third-Order Recurrence Plots

Martijn Boussé; Philippe Dreesen; Pietro Bonizzi; Joël Karel; Ralf Peeters

Time-Series Analysis Using Third-Order Recurrence Plots

Martijn Boussé, Philippe Dreesen, Pietro Bonizzi, Joël Karel, Ralf Peeters

Research output: Contribution to conference › Abstract › Academic

18 Downloads (Pure)

Abstract

Higher-order recurrence plots may enable us to reveal more structure than what is possible with traditional recurrence plots (RPs). While RPs attempt to detect recurrence relations by pair-wise comparison of time-delayed embeddings, given a time series, higher-order RPs may detect recurrences by comparing multiple time-delayed embeddings simultaneously. In this work, we limit ourselves to third-order recurrence plots (TORPs) for time series analysis, as they can still be graphed straightforwardly, and propose future directions.

Original language	English
Publication status	Published - 22 Mar 2023
Event	42nd Benelux Meeting on Systems and Control - Hotel Mennorode, Elspeet, Netherlands Duration: 21 Mar 2023 → 23 Mar 2023 https://www.beneluxmeeting.nl/2023/

Conference

Conference	42nd Benelux Meeting on Systems and Control
Abbreviated title	BMSC
Country/Territory	Netherlands
City	Elspeet
Period	21/03/23 → 23/03/23
Internet address	https://www.beneluxmeeting.nl/2023/

Keywords

recurrence plots
Tensors
time series analysis

Access to Document

bmsc2023mbousseFinal published version, 263 KB

Cite this

@conference{4b06d3ca5e0948c9a43854cb861dafa1,

title = "Time-Series Analysis Using Third-Order Recurrence Plots",

abstract = "Higher-order recurrence plots may enable us to reveal more structure than what is possible with traditional recurrence plots (RPs). While RPs attempt to detect recurrence relations by pair-wise comparison of time-delayed embeddings, given a time series, higher-order RPs may detect recurrences by comparing multiple time-delayed embeddings simultaneously. In this work, we limit ourselves to third-order recurrence plots (TORPs) for time series analysis, as they can still be graphed straightforwardly, and propose future directions.",

keywords = "recurrence plots, Tensors, time series analysis",

author = "Martijn Bouss{\'e} and Philippe Dreesen and Pietro Bonizzi and Jo{\"e}l Karel and Ralf Peeters",

year = "2023",

month = mar,

day = "22",

language = "English",

note = "42nd Benelux Meeting on Systems and Control, BMSC ; Conference date: 21-03-2023 Through 23-03-2023",

url = "https://www.beneluxmeeting.nl/2023/",

}

TY - CONF

T1 - Time-Series Analysis Using Third-Order Recurrence Plots

AU - Boussé, Martijn

AU - Dreesen, Philippe

AU - Bonizzi, Pietro

AU - Karel, Joël

AU - Peeters, Ralf

PY - 2023/3/22

Y1 - 2023/3/22

N2 - Higher-order recurrence plots may enable us to reveal more structure than what is possible with traditional recurrence plots (RPs). While RPs attempt to detect recurrence relations by pair-wise comparison of time-delayed embeddings, given a time series, higher-order RPs may detect recurrences by comparing multiple time-delayed embeddings simultaneously. In this work, we limit ourselves to third-order recurrence plots (TORPs) for time series analysis, as they can still be graphed straightforwardly, and propose future directions.

AB - Higher-order recurrence plots may enable us to reveal more structure than what is possible with traditional recurrence plots (RPs). While RPs attempt to detect recurrence relations by pair-wise comparison of time-delayed embeddings, given a time series, higher-order RPs may detect recurrences by comparing multiple time-delayed embeddings simultaneously. In this work, we limit ourselves to third-order recurrence plots (TORPs) for time series analysis, as they can still be graphed straightforwardly, and propose future directions.

KW - recurrence plots

KW - Tensors

KW - time series analysis

M3 - Abstract

T2 - 42nd Benelux Meeting on Systems and Control

Y2 - 21 March 2023 through 23 March 2023

ER -