Multiscale limited penetrable horizontal visibility graph for analyzing nonlinear time series.

Area of Science:

• Complex systems analysis
• Network science
• Time series analysis

Background:

• Visibility graphs are effective for time series analysis.
• Analyzing nonlinear time series from complex systems like EEG and two-phase flow presents challenges.
• Existing methods may not capture multiscale dynamics effectively.

Purpose of the Study:

• To develop a novel multiscale limited penetrable horizontal visibility graph (MLPHVG) for nonlinear time series analysis.
• To demonstrate the MLPHVG method's effectiveness on EEG and two-phase flow data.
• To apply MLPHVG for seizure detection and flow pattern characterization.

Main Methods:

• Development of the multiscale limited penetrable horizontal visibility graph (MLPHVG).
• Application of MLPHVG to electroencephalogram (EEG) signals and oil-water two-phase flow data.
• Integration of MLPHVG with support vector machine (SVM) for classification tasks.

Main Results:

• 100% classification accuracy in detecting epileptic seizures from EEG signals using MLPHVG and SVM.
• Successful identification and characterization of three typical oil-water flow patterns using MLPHVG.
• Demonstration of MLPHVG's capability to analyze nonlinear time series across different scales.

Conclusions:

• The MLPHVG method is a powerful tool for analyzing nonlinear time series from complex systems.
• MLPHVG offers high accuracy in biomedical signal analysis (e.g., seizure detection).
• MLPHVG effectively characterizes dynamic patterns in fluid flow, highlighting its versatility.