Abstract
Applied in many areas, from original hydrology to modern computer networking, Hurst exponent provides us with an indicator that the analyzed data is a completely random process or has underlying trends. But a good estimation of Hurst exponent remains complicated as R/S algorithm shows. Recurring to fractal mathematics, especially the research on fractal Brownian motion (fBm), wavelet packet transform is introduced to estimate Hurst exponent. Compared with wavelet transform and other estimating methods, the wavelet packet algorithm is found able to provide more accurate result. And another advantage of wavelet packet is the extensive choice of available analyzing wavelet filter functions.
| Original language | English |
|---|---|
| Title of host publication | 2006 7th International Conference on Computer-Aided Industrial Design and Conceptual Design, CAIDC |
| DOIs | |
| Publication status | Published - 2006 |
| Event | 2006 7th International Conference on Computer-Aided Industrial Design and Conceptual Design, CAIDC - Hangzhou, China Duration: 17 Nov 2006 → 19 Nov 2006 |
Publication series
| Name | 2006 7th International Conference on Computer-Aided Industrial Design and Conceptual Design, CAIDC |
|---|
Conference
| Conference | 2006 7th International Conference on Computer-Aided Industrial Design and Conceptual Design, CAIDC |
|---|---|
| Country/Territory | China |
| City | Hangzhou |
| Period | 17/11/06 → 19/11/06 |
Keywords
- FBm
- Hurst exponent
- R/S
- Wavelet packet
Fingerprint
Dive into the research topics of 'Estimating hurst exponent with wavelet packet'. Together they form a unique fingerprint.Cite this
Wang, Z., Guo, D., Li, X., & Fei, Y. (2006). Estimating hurst exponent with wavelet packet. In 2006 7th International Conference on Computer-Aided Industrial Design and Conceptual Design, CAIDC Article 4127049 (2006 7th International Conference on Computer-Aided Industrial Design and Conceptual Design, CAIDC). https://doi.org/10.1109/CAIDCD.2006.329420