A Hybrid Method Using HAVOK Analysis and Machine Learning for Predicting Chaotic Time Series

Yang, Jinhui; Zhao, Juan; Song, Junqiang; Wu, Jianping; Zhao, Cheng-Wu; Leng, Hongze

doi:10.3390/e24030408

Cited by 10 publications

(5 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The above researches have shown that the detection performance of weak pulse signals depends on the accuracy of chaotic signal prediction. The higher prediction accuracy, the better detection performance of the model [11]. Recently, methods such as fractional maximum correlation entropy algorithm, deep learning and extreme learning machine have been used to the prediction of chaotic time series [12][13][14], which significantly improves the prediction accuracy.…”

Section: Introductionmentioning

confidence: 99%

Detection of Weak Pulse Signal in Chaotic Noise - Based on Att-CNN-LSTM model

Shen,

Zhao,

Liu

et al. 2024

Preprint

View full text Add to dashboard Cite

A hybrid neural network based on the attention mechanism was proposed to achieve the detection of weak pulse signals in chaotic noise. Firstly, based on the high sensitivity to small interference and short-term predictability of chaotic signals , the phase space of observed signals was reconstruction. Then, Att-CNN-LSTM, a hybrid neural network based on the attention mechanism was proposed to predict chaotic signals, and the one-step prediction error was obtained, The detection problem of the observed signal can be transformed into a signal detection problem for the one-step prediction error. Finally, the weak impulse signal was detected from the prediction error by using the Z-test method. In the simulation experiments, the results of the proposed model were compared with those of the single convolutional neural network (CNN) and long short-term memory neural network (LSTM) model, the least square support vector machine, and the CNN-LSTM model without the attention mechanism. The simulation results show that the proposed model has higher prediction accuracy than other models at different signal-to-noise ratios(SNR), and achieves good performance in the detection of weak pulse signals when the SNR is greater than -140.91dB.

show abstract

Section: Introductionmentioning

confidence: 99%

Detection of Weak Pulse Signal in Chaotic Noise - Based on Att-CNN-LSTM model

Shen,

Zhao,

Liu

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…Karimov et al constructed a chaotic circuit from data to identify chaos systems [14]. Yang et al introduced the Hankel Alternative View of Koopman analysis to decompose chaotic dynamics into a linear model with intermittent forcing [15]. The most widely used methods of chaotic time series analysis are neural network-related methods, which are classified into artificial neural networks (ANN) [16][17][18][19][20], fuzzy neural networks (FNN) [21][22][23][24], optimization algorithms with ANN [25][26][27][28], and wavelet neural networks (WNN) [29][30][31][32].…”

Section: Introductionmentioning

confidence: 99%

Multilayer Perceptron Network Optimization for Chaotic Time Series Modeling

Qiao

Liang

Tavares

et al. 2023

Entropy

View full text Add to dashboard Cite

Chaotic time series are widely present in practice, but due to their characteristics—such as internal randomness, nonlinearity, and long-term unpredictability—it is difficult to achieve high-precision intermediate or long-term predictions. Multi-layer perceptron (MLP) networks are an effective tool for chaotic time series modeling. Focusing on chaotic time series modeling, this paper presents a generalized degree of freedom approximation method of MLP. We then obtain its Akachi information criterion, which is designed as the loss function for training, hence developing an overall framework for chaotic time series analysis, including phase space reconstruction, model training, and model selection. To verify the effectiveness of the proposed method, it is applied to two artificial chaotic time series and two real-world chaotic time series. The numerical results show that the proposed optimized method is effective to obtain the best model from a group of candidates. Moreover, the optimized models perform very well in multi-step prediction tasks.

show abstract

“…Recently, machine learning has accomplished extraordinary triumphs in the avenue of computer vision [1], semantic segmentation [2], regression prediction [3], natural language processing [4], etc. However, two problems of traditional machine learning are gradually exposed: Firstly, traditional machine learning requires a large amount of labeled data, and the cost of collecting and labeling data is expensive; thus, it is difficult to be applied in fields that lack the data required for training models.…”

Section: Introductionmentioning

confidence: 99%

Research on Adversarial Domain Adaptation Method and Its Application in Power Load Forecasting

Huang

Yin

2022

Mathematics

View full text Add to dashboard Cite

Domain adaptation has been used to transfer the knowledge from the source domain to the target domain where training data is insufficient in the target domain; thus, it can overcome the data shortage problem of power load forecasting effectively. Inspired by Generative Adversarial Networks (GANs), adversarial domain adaptation transfers knowledge in adversarial learning. Existing adversarial domain adaptation faces the problems of adversarial disequilibrium and a lack of transferability quantification, which will eventually decrease the prediction accuracy. To address this issue, a novel adversarial domain adaptation method is proposed. Firstly, by analyzing the causes of the adversarial disequilibrium, an initial state fusion strategy is proposed to improve the reliability of the domain discriminator, thus maintaining the adversarial equilibrium. Secondly, domain similarity is calculated to quantify the transferability of source domain samples based on information entropy; through weighting in the process of domain alignment, the knowledge is transferred selectively and the negative transfer is suppressed. Finally, the Building Data Genome Project 2 (BDGP2) dataset is used to validate the proposed method. The experimental results demonstrate that the proposed method can alleviate the problem of adversarial disequilibrium and reasonably quantify the transferability to improve the accuracy of power load forecasting.

show abstract

A Hybrid Method Using HAVOK Analysis and Machine Learning for Predicting Chaotic Time Series

Cited by 10 publications

References 29 publications

Detection of Weak Pulse Signal in Chaotic Noise - Based on Att-CNN-LSTM model

Detection of Weak Pulse Signal in Chaotic Noise - Based on Att-CNN-LSTM model

Multilayer Perceptron Network Optimization for Chaotic Time Series Modeling

Research on Adversarial Domain Adaptation Method and Its Application in Power Load Forecasting

Contact Info

Product

Resources

About