Related Experiment Videos
An adaptive way for improving noise reduction using local geometric projection.
Alexandros Leontitsis1, Tassos Bountis, Jenny Pagge
1Department of Education, University of Ioannina, 45110-Dourouti, Ioannina, Greece. rne00743@cc.uoi.gr
Chaos (Woodbury, N.Y.)
|March 9, 2004
Summary
This study introduces adaptive local geometric projection for improved noise reduction in chaotic time series. The method effectively separates signal from noise by analyzing neighborhood characteristics for more accurate data analysis.
Area of Science:
- Data analysis and signal processing
- Nonlinear dynamics and chaos theory
Background:
- Noise reduction is crucial for accurate analysis of complex datasets.
- Traditional methods may struggle with the unique characteristics of chaotic time series.
Purpose of the Study:
- To develop an adaptive noise reduction technique using local geometric projection.
- To enhance the accuracy of signal extraction from noisy chaotic data.
Main Methods:
- Orthogonal projection of phase space points onto identified subspaces.
- Eigendirection analysis to distinguish signal and noise subspaces within local neighborhoods.
- A criterion based on logarithmic differences of eigendirection lengths for subspace separation.
Main Results:
- Demonstrated successful noise reduction on chaotic time series (Henon map, Ikeda map).
- Validated the technique on a real-world financial dataset (Nasdaq Composite index).
- The adaptive approach effectively captures local data characteristics for improved noise filtering.
Conclusions:
- Adaptive local geometric projection offers a robust method for noise reduction in chaotic systems.
- The proposed criterion provides an effective way to differentiate signal from noise subspaces.
- This technique has broad applicability in analyzing complex and noisy time series data.