A Long-Short Term Memory Network for Chaotic Time Series Prediction

Margarita Terziyska; Zhelyazko  Terziyski; Yancho Todorov

doi:10.1109/BdKCSE53180.2021.9627283

A Long-Short Term Memory Network for Chaotic Time Series Prediction

Margarita Terziyska, Zhelyazko Terziyski, Yancho Todorov

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

1 Citation (Scopus)

Abstract

This paper demonstrates the development of an Long-Short Term Memory (LSTM) network and its application for predicting Chaotic Time Series (CTS). LSTM is a Deep Learning (DL) architecture and a type of Recurrent Neural Network (RNN). Its ability to memorise past inputs make it extremely suitable for studying data sequences, such as CTS. To test the performance of a developed LSTM, several simulation studies were realized in a Matlab environment. The developed DL architecture was used to predict well-known standard CTSs, namely Mackey-Glass (MG), Rössler, and Lorenz. The obtained results were compared with Adaptive N euro Fuzzy Inference system (ANFIS) and Distributed Adaptive Neuro-Fuzzy Architecture (DANF A), previously developed by the authors. The comparison is made on the basis of Root Mean Squared Error (RMSE). It was found that the proposed LSTM structure is able to generalize the generated series and it has higher accuracy than the other two models.

Original language	English
Title of host publication	Big Data, Knowledge and Control Systems Engineering - Proceedings of the 7th International Conference, BdKCSE 2021
Editors	Rumen Andreev, Lyubka Doukovska, Svetozar Ilchev
Publisher	IEEE Institute of Electrical and Electronic Engineers
ISBN (Electronic)	978-1-6654-1042-7
ISBN (Print)	978-1-6654-1043-4
DOIs	https://doi.org/10.1109/BdKCSE53180.2021.9627283
Publication status	Published - 28 Oct 2021
MoE publication type	A4 Article in a conference publication
Event	2021 Big Data, Knowledge and Control Systems Engineering, BdKCSE - Sofia, Bulgaria Duration: 28 Oct 2021 → 29 Oct 2021

Conference

Conference	2021 Big Data, Knowledge and Control Systems Engineering, BdKCSE
Abbreviated title	BdKCSE
Country/Territory	Bulgaria
City	Sofia
Period	28/10/21 → 29/10/21

Keywords

chaotic time series
deep learning
LSTM
prediction

Access to Document

10.1109/BdKCSE53180.2021.9627283

Cite this

Terziyska, M., Terziyski, Z., & Todorov, Y. (2021). A Long-Short Term Memory Network for Chaotic Time Series Prediction. In R. Andreev, L. Doukovska, & S. Ilchev (Eds.), Big Data, Knowledge and Control Systems Engineering - Proceedings of the 7th International Conference, BdKCSE 2021 IEEE Institute of Electrical and Electronic Engineers. https://doi.org/10.1109/BdKCSE53180.2021.9627283

Terziyska, Margarita ; Terziyski, Zhelyazko ; Todorov, Yancho. / A Long-Short Term Memory Network for Chaotic Time Series Prediction. Big Data, Knowledge and Control Systems Engineering - Proceedings of the 7th International Conference, BdKCSE 2021. editor / Rumen Andreev ; Lyubka Doukovska ; Svetozar Ilchev. IEEE Institute of Electrical and Electronic Engineers, 2021.

@inproceedings{1ab5a873a43e48a091ad4c1ada5d8f13,

title = "A Long-Short Term Memory Network for Chaotic Time Series Prediction",

abstract = "This paper demonstrates the development of an Long-Short Term Memory (LSTM) network and its application for predicting Chaotic Time Series (CTS). LSTM is a Deep Learning (DL) architecture and a type of Recurrent Neural Network (RNN). Its ability to memorise past inputs make it extremely suitable for studying data sequences, such as CTS. To test the performance of a developed LSTM, several simulation studies were realized in a Matlab environment. The developed DL architecture was used to predict well-known standard CTSs, namely Mackey-Glass (MG), R{\"o}ssler, and Lorenz. The obtained results were compared with Adaptive N euro Fuzzy Inference system (ANFIS) and Distributed Adaptive Neuro-Fuzzy Architecture (DANF A), previously developed by the authors. The comparison is made on the basis of Root Mean Squared Error (RMSE). It was found that the proposed LSTM structure is able to generalize the generated series and it has higher accuracy than the other two models.",

keywords = "chaotic time series, deep learning, LSTM, prediction",

author = "Margarita Terziyska and Zhelyazko Terziyski and Yancho Todorov",

note = "Funding Information: The authors would like to acknowledge the support of the Research and International Sector at fhe University of Food Technologies (UFT), Plovdiv, Bulgaria under research project N, 08/21-H funded through the UFT's Science Fund. Publisher Copyright: {\textcopyright} 2021 IEEE.; 2021 Big Data, Knowledge and Control Systems Engineering, BdKCSE, BdKCSE ; Conference date: 28-10-2021 Through 29-10-2021",

year = "2021",

month = oct,

day = "28",

doi = "10.1109/BdKCSE53180.2021.9627283",

language = "English",

isbn = "978-1-6654-1043-4",

editor = "Rumen Andreev and Lyubka Doukovska and Svetozar Ilchev",

booktitle = "Big Data, Knowledge and Control Systems Engineering - Proceedings of the 7th International Conference, BdKCSE 2021",

publisher = "IEEE Institute of Electrical and Electronic Engineers",

address = "United States",

}

Terziyska, M, Terziyski, Z & Todorov, Y 2021, A Long-Short Term Memory Network for Chaotic Time Series Prediction. in R Andreev, L Doukovska & S Ilchev (eds), Big Data, Knowledge and Control Systems Engineering - Proceedings of the 7th International Conference, BdKCSE 2021. IEEE Institute of Electrical and Electronic Engineers, 2021 Big Data, Knowledge and Control Systems Engineering, BdKCSE, Sofia, Bulgaria, 28/10/21. https://doi.org/10.1109/BdKCSE53180.2021.9627283

A Long-Short Term Memory Network for Chaotic Time Series Prediction. / Terziyska, Margarita; Terziyski, Zhelyazko ; Todorov, Yancho.
Big Data, Knowledge and Control Systems Engineering - Proceedings of the 7th International Conference, BdKCSE 2021. ed. / Rumen Andreev; Lyubka Doukovska; Svetozar Ilchev. IEEE Institute of Electrical and Electronic Engineers, 2021.

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

TY - GEN

T1 - A Long-Short Term Memory Network for Chaotic Time Series Prediction

AU - Terziyska, Margarita

AU - Terziyski, Zhelyazko

AU - Todorov, Yancho

N1 - Funding Information: The authors would like to acknowledge the support of the Research and International Sector at fhe University of Food Technologies (UFT), Plovdiv, Bulgaria under research project N, 08/21-H funded through the UFT's Science Fund. Publisher Copyright: © 2021 IEEE.

PY - 2021/10/28

Y1 - 2021/10/28

N2 - This paper demonstrates the development of an Long-Short Term Memory (LSTM) network and its application for predicting Chaotic Time Series (CTS). LSTM is a Deep Learning (DL) architecture and a type of Recurrent Neural Network (RNN). Its ability to memorise past inputs make it extremely suitable for studying data sequences, such as CTS. To test the performance of a developed LSTM, several simulation studies were realized in a Matlab environment. The developed DL architecture was used to predict well-known standard CTSs, namely Mackey-Glass (MG), Rössler, and Lorenz. The obtained results were compared with Adaptive N euro Fuzzy Inference system (ANFIS) and Distributed Adaptive Neuro-Fuzzy Architecture (DANF A), previously developed by the authors. The comparison is made on the basis of Root Mean Squared Error (RMSE). It was found that the proposed LSTM structure is able to generalize the generated series and it has higher accuracy than the other two models.

AB - This paper demonstrates the development of an Long-Short Term Memory (LSTM) network and its application for predicting Chaotic Time Series (CTS). LSTM is a Deep Learning (DL) architecture and a type of Recurrent Neural Network (RNN). Its ability to memorise past inputs make it extremely suitable for studying data sequences, such as CTS. To test the performance of a developed LSTM, several simulation studies were realized in a Matlab environment. The developed DL architecture was used to predict well-known standard CTSs, namely Mackey-Glass (MG), Rössler, and Lorenz. The obtained results were compared with Adaptive N euro Fuzzy Inference system (ANFIS) and Distributed Adaptive Neuro-Fuzzy Architecture (DANF A), previously developed by the authors. The comparison is made on the basis of Root Mean Squared Error (RMSE). It was found that the proposed LSTM structure is able to generalize the generated series and it has higher accuracy than the other two models.

KW - chaotic time series

KW - deep learning

KW - LSTM

KW - prediction

UR - http://www.scopus.com/inward/record.url?scp=85123995035&partnerID=8YFLogxK

U2 - 10.1109/BdKCSE53180.2021.9627283

DO - 10.1109/BdKCSE53180.2021.9627283

M3 - Conference article in proceedings

SN - 978-1-6654-1043-4

BT - Big Data, Knowledge and Control Systems Engineering - Proceedings of the 7th International Conference, BdKCSE 2021

A2 - Andreev, Rumen

A2 - Doukovska, Lyubka

A2 - Ilchev, Svetozar

PB - IEEE Institute of Electrical and Electronic Engineers

T2 - 2021 Big Data, Knowledge and Control Systems Engineering, BdKCSE

Y2 - 28 October 2021 through 29 October 2021

ER -

A Long-Short Term Memory Network for Chaotic Time Series Prediction

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this