Multi-scale transition network approaches for nonlinear time series analysis

Wang, Xiaoyan; Han, Xiujing; Zhang-Yao, Chen; Bi, Qinsheng; Guan, Shuguang; Zou, Yong

doi:10.1016/j.chaos.2022.112026

Cited by 22 publications

(4 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The electrocardiogram (ECG) is a fundamental diagnostic tool used in medicine to monitor and evaluate the electrical activity of the heart [ [1] , [2] , [3] ]. It records the electrical impulses generated by the heart during each heartbeat cycle.…”

Section: Introductionmentioning

confidence: 99%

Robust QRS detection based on simulated degenerate optical parametric oscillator-assisted neural network

Liao,

Shi,

Sarker

et al. 2024

Heliyon

View full text Add to dashboard Cite

Section: Introductionmentioning

confidence: 99%

Robust QRS detection based on simulated degenerate optical parametric oscillator-assisted neural network

Liao,

Shi,

Sarker

et al. 2024

Heliyon

View full text Add to dashboard Cite

“…The scrutiny of time series data has garnered considerable attention, propelled by the necessity to decipher meaningful insights from complex temporal patterns [4,5]. RNNs have surfaced as a cornerstone in this domain, facilitating the modeling of sequential dependencies across varied applications, from natural language processing to financial forecasting [6].…”

Section: Introductionmentioning

confidence: 99%

Multi-Variable Time Series Decoding with Long Short-Term Memory and Mixture Attention

Seddik,

Routaib,

Elhaddadi

2023

ATAIML

View full text Add to dashboard Cite

The task of interpreting multi-variable time series data, while also forecasting outcomes accurately, is an ongoing challenge within the machine learning domain. This study presents an advanced method of utilizing Long Short-Term Memory (LSTM) recurrent neural networks in the analysis of such data, with specific attention to both target and exogenous variables. The novel approach aims to extract hidden states that are unique to individual variables, thereby capturing the distinctive dynamics inherent in multi-variable time series and allowing the elucidation of each variable's contribution to predictive outcomes. A pioneering mixture attention mechanism is introduced, which, by leveraging the aforementioned variable-specific hidden states, characterizes the generative process of the target variable. The study further enhances this methodology by formulating associated training techniques that permit concurrent learning of network parameters, variable interactions, and temporal significance with respect to the target prediction. The effectiveness of this approach is empirically validated through rigorous experimentation on three real-world datasets, including the 2022 closing prices of three major stocks -Apple (AAPL), Amazon (AMZN), and Microsoft (MSFT). The results demonstrated superior predictive performance, attributable to the successful encapsulation of the diverse dynamics of different variables. Furthermore, the study provides a comprehensive evaluation of the interpretability outcomes, both qualitatively and quantitatively. The presented framework thus holds substantial promise as a comprehensive solution that not only enhances prediction accuracy but also aids in the extraction of valuable insights from complex multi-variable datasets.

show abstract

“…The versatility of this method extends to various applications, such as distinguishing between different consciousness states [15], analyzing EEG data [16], investigating stock markets [17,18], and examining transportation data [13]. In combination with complex network techniques for nonlinear time series analysis, such as visibility or recurrence networks [10,[19][20][21][22][23], this approach presents a valuable addition to the set of available tools.…”

Section: Introductionmentioning

confidence: 99%

Unveiling the Connectivity of Complex Networks Using Ordinal Transition Methods

Almendral

Leyva

Sendiña‐Nadal

2023

Entropy

View full text Add to dashboard Cite

Ordinal measures provide a valuable collection of tools for analyzing correlated data series. However, using these methods to understand information interchange in the networks of dynamical systems, and uncover the interplay between dynamics and structure during the synchronization process, remains relatively unexplored. Here, we compare the ordinal permutation entropy, a standard complexity measure in the literature, and the permutation entropy of the ordinal transition probability matrix that describes the transitions between the ordinal patterns derived from a time series. We find that the permutation entropy based on the ordinal transition matrix outperforms the rest of the tested measures in discriminating the topological role of networked chaotic Rössler systems. Since the method is based on permutation entropy measures, it can be applied to arbitrary real-world time series exhibiting correlations originating from an existing underlying unknown network structure. In particular, we show the effectiveness of our method using experimental datasets of networks of nonlinear oscillators.

show abstract

Multi-scale transition network approaches for nonlinear time series analysis

Cited by 22 publications

References 24 publications

Robust QRS detection based on simulated degenerate optical parametric oscillator-assisted neural network

Robust QRS detection based on simulated degenerate optical parametric oscillator-assisted neural network

Multi-Variable Time Series Decoding with Long Short-Term Memory and Mixture Attention

Unveiling the Connectivity of Complex Networks Using Ordinal Transition Methods

Contact Info

Product

Resources

About