Peak Detection and Correlation Analysis in Noisy Time Series Data
This paper focus on, the study of correlation (dependency) between the extreme trends (peaks) in multi-variant noise time series data, In some sense, the extreme events disrupt the underlying structure distribution in the data. The peaks are identified using a data-driven algorithm. It is observed that all majority of peak locations are identified using this method. We also evaluate its robustness by giving the different size of data records.
KeywordsPeaks Correlation Buoy Multi-variant time series
This work was completed in INCOIS Hyderabad. Authors wish to thank Director INCOIS, Hyderabad for the encouragement and facilities provided. Authors also acknowledge the support and guidance of other INCOIS scientists throughout working on this project and preparing this manuscript. We would also like to express our gratitude to Prof. S.C. Satapathy (Head of Department), ANITS, Visakhapatnam for his continuous support and encouragement.
- 1.L.M Bhar and V.K Sharma, “Time Series Analysis”, Indian agriculture statistics research institute, New Delhi.Google Scholar
- 2.Roger Schneider, “Survey of Peaks/Valleys identification in Time Series”, August 23, 2011.Google Scholar
- 3.Girish Palshikar, “Simple Algorithms for Peak Detection in Time-Series”, article, Jan 2009.Google Scholar
- 4.Sayanti Chattopadhyay, Susmita Das, “Design and Simulation Approach Introduced to ECG Peak Detection with study on different cardiovascular Diseases”, IJSRP, Vol. 2, December 2012.Google Scholar
- 7.Felix Scholkmann, Jens Boss and Martin Wolf, “An Efficient Algorithm for Automatic Peak Detection in Noisy Periodic and Quasi-Periodic Signals”, article, 2012.Google Scholar
- 8.Julian D. Olden, Bryan D. Neff, “Cross correlation bias in lag analysis of aquatic time series”, Springer Marine Biology pp. 1063–1070, 2001.Google Scholar