A preliminary survey of analyzing dynamic time-varying financial networks using graph kernels

Lixin Cui, Lu Bai*, Luca Rossi, Zhihong Zhang, Yuhang Jiao, Edwin R. Hancock

*Corresponding author for this work

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

Abstract

In this paper, we investigate whether graph kernels can be used as a means of analyzing time-varying financial market networks. Specifically, we aim to identify the significant financial incident that changes the financial network properties through graph kernels. Our financial networks are abstracted from the New York Stock Exchange (NYSE) data over 6004 trading days, where each vertex represents the individual daily return price time series of a stock and each edge represents the correlation between pairwise series. We propose to use two state-of-the-art graph kernels for the analysis, i.e., the Jensen-Shannon graph kernel and the Weisfeiler-Lehman subtree kernel. The reason of using the two kernels is that they are the representative methods of global graph kernels and local graph kernels, respectively. We perform kernel Principle Components Analysis (kPCA) associated with each kernel matrix to embed the networks into a 3-dimensional principle space, where the time-varying networks of all trading days are visualized. Experimental results on the financial time series of NYSE dataset demonstrate that graph kernels can well distinguish abrupt changes of financial networks with time, and provide a more effective alternative way of analyzing original multiple co-evolving financial time series. We theoretically indicate the perspective of developing novel graph kernels on time-varying networks for multiple co-evolving time series analysis in future work.

Original languageEnglish
Title of host publicationStructural, Syntactic, and Statistical Pattern Recognition - Joint IAPR International Workshop, S+SSPR 2018, Proceedings
PublisherSpringer
Pages237-247
Number of pages11
ISBN (Electronic)978-3-319-97785-0
ISBN (Print)9783319977843
DOIs
Publication statusPublished - 2 Aug 2018
EventJoint IAPR International Workshops on Structural and Syntactic Pattern Recognition, SSPR 2018 and Statistical Techniques in Pattern Recognition, SPR 2018 - Beijing, China
Duration: 17 Aug 201819 Aug 2018

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11004 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

ConferenceJoint IAPR International Workshops on Structural and Syntactic Pattern Recognition, SSPR 2018 and Statistical Techniques in Pattern Recognition, SPR 2018
CountryChina
CityBeijing
Period17/08/1819/08/18

Fingerprint

Time varying networks
Time series
Time-varying
kernel
Graph in graph theory
Time series analysis
Electronic data interchange
Financial Time Series
Principle Component Analysis
Data Exchange
Time Series Analysis
Financial Markets
Pairwise

Keywords

  • Graph kernels
  • NYSE dataset
  • Time-varying financial networks

Cite this

Cui, L., Bai, L., Rossi, L., Zhang, Z., Jiao, Y., & Hancock, E. R. (2018). A preliminary survey of analyzing dynamic time-varying financial networks using graph kernels. In Structural, Syntactic, and Statistical Pattern Recognition - Joint IAPR International Workshop, S+SSPR 2018, Proceedings (pp. 237-247). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11004 LNCS). Springer. https://doi.org/10.1007/978-3-319-97785-0_23
Cui, Lixin ; Bai, Lu ; Rossi, Luca ; Zhang, Zhihong ; Jiao, Yuhang ; Hancock, Edwin R. / A preliminary survey of analyzing dynamic time-varying financial networks using graph kernels. Structural, Syntactic, and Statistical Pattern Recognition - Joint IAPR International Workshop, S+SSPR 2018, Proceedings. Springer, 2018. pp. 237-247 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{d252e8f8cbb744a9a15316f1211be02e,
title = "A preliminary survey of analyzing dynamic time-varying financial networks using graph kernels",
abstract = "In this paper, we investigate whether graph kernels can be used as a means of analyzing time-varying financial market networks. Specifically, we aim to identify the significant financial incident that changes the financial network properties through graph kernels. Our financial networks are abstracted from the New York Stock Exchange (NYSE) data over 6004 trading days, where each vertex represents the individual daily return price time series of a stock and each edge represents the correlation between pairwise series. We propose to use two state-of-the-art graph kernels for the analysis, i.e., the Jensen-Shannon graph kernel and the Weisfeiler-Lehman subtree kernel. The reason of using the two kernels is that they are the representative methods of global graph kernels and local graph kernels, respectively. We perform kernel Principle Components Analysis (kPCA) associated with each kernel matrix to embed the networks into a 3-dimensional principle space, where the time-varying networks of all trading days are visualized. Experimental results on the financial time series of NYSE dataset demonstrate that graph kernels can well distinguish abrupt changes of financial networks with time, and provide a more effective alternative way of analyzing original multiple co-evolving financial time series. We theoretically indicate the perspective of developing novel graph kernels on time-varying networks for multiple co-evolving time series analysis in future work.",
keywords = "Graph kernels, NYSE dataset, Time-varying financial networks",
author = "Lixin Cui and Lu Bai and Luca Rossi and Zhihong Zhang and Yuhang Jiao and Hancock, {Edwin R.}",
year = "2018",
month = "8",
day = "2",
doi = "10.1007/978-3-319-97785-0_23",
language = "English",
isbn = "9783319977843",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer",
pages = "237--247",
booktitle = "Structural, Syntactic, and Statistical Pattern Recognition - Joint IAPR International Workshop, S+SSPR 2018, Proceedings",
address = "Germany",

}

Cui, L, Bai, L, Rossi, L, Zhang, Z, Jiao, Y & Hancock, ER 2018, A preliminary survey of analyzing dynamic time-varying financial networks using graph kernels. in Structural, Syntactic, and Statistical Pattern Recognition - Joint IAPR International Workshop, S+SSPR 2018, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11004 LNCS, Springer, pp. 237-247, Joint IAPR International Workshops on Structural and Syntactic Pattern Recognition, SSPR 2018 and Statistical Techniques in Pattern Recognition, SPR 2018, Beijing, China, 17/08/18. https://doi.org/10.1007/978-3-319-97785-0_23

A preliminary survey of analyzing dynamic time-varying financial networks using graph kernels. / Cui, Lixin; Bai, Lu; Rossi, Luca; Zhang, Zhihong; Jiao, Yuhang; Hancock, Edwin R.

Structural, Syntactic, and Statistical Pattern Recognition - Joint IAPR International Workshop, S+SSPR 2018, Proceedings. Springer, 2018. p. 237-247 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11004 LNCS).

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

TY - GEN

T1 - A preliminary survey of analyzing dynamic time-varying financial networks using graph kernels

AU - Cui, Lixin

AU - Bai, Lu

AU - Rossi, Luca

AU - Zhang, Zhihong

AU - Jiao, Yuhang

AU - Hancock, Edwin R.

PY - 2018/8/2

Y1 - 2018/8/2

N2 - In this paper, we investigate whether graph kernels can be used as a means of analyzing time-varying financial market networks. Specifically, we aim to identify the significant financial incident that changes the financial network properties through graph kernels. Our financial networks are abstracted from the New York Stock Exchange (NYSE) data over 6004 trading days, where each vertex represents the individual daily return price time series of a stock and each edge represents the correlation between pairwise series. We propose to use two state-of-the-art graph kernels for the analysis, i.e., the Jensen-Shannon graph kernel and the Weisfeiler-Lehman subtree kernel. The reason of using the two kernels is that they are the representative methods of global graph kernels and local graph kernels, respectively. We perform kernel Principle Components Analysis (kPCA) associated with each kernel matrix to embed the networks into a 3-dimensional principle space, where the time-varying networks of all trading days are visualized. Experimental results on the financial time series of NYSE dataset demonstrate that graph kernels can well distinguish abrupt changes of financial networks with time, and provide a more effective alternative way of analyzing original multiple co-evolving financial time series. We theoretically indicate the perspective of developing novel graph kernels on time-varying networks for multiple co-evolving time series analysis in future work.

AB - In this paper, we investigate whether graph kernels can be used as a means of analyzing time-varying financial market networks. Specifically, we aim to identify the significant financial incident that changes the financial network properties through graph kernels. Our financial networks are abstracted from the New York Stock Exchange (NYSE) data over 6004 trading days, where each vertex represents the individual daily return price time series of a stock and each edge represents the correlation between pairwise series. We propose to use two state-of-the-art graph kernels for the analysis, i.e., the Jensen-Shannon graph kernel and the Weisfeiler-Lehman subtree kernel. The reason of using the two kernels is that they are the representative methods of global graph kernels and local graph kernels, respectively. We perform kernel Principle Components Analysis (kPCA) associated with each kernel matrix to embed the networks into a 3-dimensional principle space, where the time-varying networks of all trading days are visualized. Experimental results on the financial time series of NYSE dataset demonstrate that graph kernels can well distinguish abrupt changes of financial networks with time, and provide a more effective alternative way of analyzing original multiple co-evolving financial time series. We theoretically indicate the perspective of developing novel graph kernels on time-varying networks for multiple co-evolving time series analysis in future work.

KW - Graph kernels

KW - NYSE dataset

KW - Time-varying financial networks

UR - http://www.scopus.com/inward/record.url?scp=85052215974&partnerID=8YFLogxK

UR - https://link.springer.com/chapter/10.1007%2F978-3-319-97785-0_23

U2 - 10.1007/978-3-319-97785-0_23

DO - 10.1007/978-3-319-97785-0_23

M3 - Conference contribution

AN - SCOPUS:85052215974

SN - 9783319977843

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 237

EP - 247

BT - Structural, Syntactic, and Statistical Pattern Recognition - Joint IAPR International Workshop, S+SSPR 2018, Proceedings

PB - Springer

ER -

Cui L, Bai L, Rossi L, Zhang Z, Jiao Y, Hancock ER. A preliminary survey of analyzing dynamic time-varying financial networks using graph kernels. In Structural, Syntactic, and Statistical Pattern Recognition - Joint IAPR International Workshop, S+SSPR 2018, Proceedings. Springer. 2018. p. 237-247. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-97785-0_23