A Novel Discrete Time Series Representation With De Bruijn Graphs for Enhanced Forecasting Using TimesNet

In this paper, we present a novel method for advancing time series forecasting by representing discretized time series data through de Bruijn Graphs (dBGs). This method harnesses the capability of dBGs to encapsulate and project future states from historical sequences, thus enhancing predictive anal...

Full description

Saved in:
Bibliographic Details
Main Authors: Mert Onur Cakiroglu, Hasan Kurban, Elham Buxton, Mehmet Dalkilic
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11079555/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850067561323954176
author Mert Onur Cakiroglu
Hasan Kurban
Elham Buxton
Mehmet Dalkilic
author_facet Mert Onur Cakiroglu
Hasan Kurban
Elham Buxton
Mehmet Dalkilic
author_sort Mert Onur Cakiroglu
collection DOAJ
description In this paper, we present a novel method for advancing time series forecasting by representing discretized time series data through de Bruijn Graphs (dBGs). This method harnesses the capability of dBGs to encapsulate and project future states from historical sequences, thus enhancing predictive analytics in time series. Our approach is multi-faceted, involving: 1) encoding time series data as a dBG; 2) the application of graph representation learning, specifically struct2vec, to distill salient features from dBG constructed from time series and 3) the seamless integration of these extracted features into the state of the art TimesNet model to bolster short-term forecasting accuracy. Empirical evaluations conducted on the M4 datasets illustrate that our approach not only maintains the intrinsic dynamics of the time series but also achieves notable improvements in forecasting performance across diverse datasets. All the code developed for this study can be found at: <uri>https://github.com/KurbanIntelligenceLab/dBGTime-Series-Library</uri>
format Article
id doaj-art-782a9b35d0664ca69dcf493d22613c4d
institution DOAJ
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-782a9b35d0664ca69dcf493d22613c4d2025-08-20T02:48:16ZengIEEEIEEE Access2169-35362025-01-011312318212319810.1109/ACCESS.2025.358850711079555A Novel Discrete Time Series Representation With De Bruijn Graphs for Enhanced Forecasting Using TimesNetMert Onur Cakiroglu0https://orcid.org/0009-0001-0798-1361Hasan Kurban1https://orcid.org/0000-0003-3142-2866Elham Buxton2https://orcid.org/0000-0001-7774-4604Mehmet Dalkilic3Department of Computer Science, Indiana University Bloomington, Bloomington, IN, USACollege of Science and Engineering, Hamad Bin Khalifa University, Doha, QatarDepartment of Computer Science, University of Illinois Springfield, Springfield, IL, USADepartment of Computer Science, Indiana University Bloomington, Bloomington, IN, USAIn this paper, we present a novel method for advancing time series forecasting by representing discretized time series data through de Bruijn Graphs (dBGs). This method harnesses the capability of dBGs to encapsulate and project future states from historical sequences, thus enhancing predictive analytics in time series. Our approach is multi-faceted, involving: 1) encoding time series data as a dBG; 2) the application of graph representation learning, specifically struct2vec, to distill salient features from dBG constructed from time series and 3) the seamless integration of these extracted features into the state of the art TimesNet model to bolster short-term forecasting accuracy. Empirical evaluations conducted on the M4 datasets illustrate that our approach not only maintains the intrinsic dynamics of the time series but also achieves notable improvements in forecasting performance across diverse datasets. All the code developed for this study can be found at: <uri>https://github.com/KurbanIntelligenceLab/dBGTime-Series-Library</uri>https://ieeexplore.ieee.org/document/11079555/Time series analysisDe Bruijn graphTimesNetgraph embeddings
spellingShingle Mert Onur Cakiroglu
Hasan Kurban
Elham Buxton
Mehmet Dalkilic
A Novel Discrete Time Series Representation With De Bruijn Graphs for Enhanced Forecasting Using TimesNet
IEEE Access
Time series analysis
De Bruijn graph
TimesNet
graph embeddings
title A Novel Discrete Time Series Representation With De Bruijn Graphs for Enhanced Forecasting Using TimesNet
title_full A Novel Discrete Time Series Representation With De Bruijn Graphs for Enhanced Forecasting Using TimesNet
title_fullStr A Novel Discrete Time Series Representation With De Bruijn Graphs for Enhanced Forecasting Using TimesNet
title_full_unstemmed A Novel Discrete Time Series Representation With De Bruijn Graphs for Enhanced Forecasting Using TimesNet
title_short A Novel Discrete Time Series Representation With De Bruijn Graphs for Enhanced Forecasting Using TimesNet
title_sort novel discrete time series representation with de bruijn graphs for enhanced forecasting using timesnet
topic Time series analysis
De Bruijn graph
TimesNet
graph embeddings
url https://ieeexplore.ieee.org/document/11079555/
work_keys_str_mv AT mertonurcakiroglu anoveldiscretetimeseriesrepresentationwithdebruijngraphsforenhancedforecastingusingtimesnet
AT hasankurban anoveldiscretetimeseriesrepresentationwithdebruijngraphsforenhancedforecastingusingtimesnet
AT elhambuxton anoveldiscretetimeseriesrepresentationwithdebruijngraphsforenhancedforecastingusingtimesnet
AT mehmetdalkilic anoveldiscretetimeseriesrepresentationwithdebruijngraphsforenhancedforecastingusingtimesnet
AT mertonurcakiroglu noveldiscretetimeseriesrepresentationwithdebruijngraphsforenhancedforecastingusingtimesnet
AT hasankurban noveldiscretetimeseriesrepresentationwithdebruijngraphsforenhancedforecastingusingtimesnet
AT elhambuxton noveldiscretetimeseriesrepresentationwithdebruijngraphsforenhancedforecastingusingtimesnet
AT mehmetdalkilic noveldiscretetimeseriesrepresentationwithdebruijngraphsforenhancedforecastingusingtimesnet