SiZer for time series: A new approach to the analysis of trends

• DOI: 10.1214/07-ejs006
• Publication Date: 2007-06-28T07:00:20Z
• Authors: Vitaliana Rondonotti; J. S. Marron; Cheolwoo Park
• Keywords: Spurious relationship; Statistician; Smoothing; Data set

Smoothing methods and SiZer are a useful statistical tool for discovering statistically significant structure in data. Based on scale space ideas originally developed the computer vision literature, (SIgnificant ZERo crossing of derivatives) is graphical device to assess which observed features `really there' just spurious sampling artifacts. In this paper, we develop like time series analysis address important issue significance trends. This not straightforward extension, since one data set does contain information needed distinguish `trend' from `dependence'. A new visualization proposed, shows statistician range trade-offs that available. Simulation real results illustrate effectiveness method.