Publication Date
12-2012
Abstract
In the analysis of time series, it is important to decompose the original values into trend, cycle, seasonal component, and noise. In this paper, we provide a theoretical justification of the fact that the F-transform can be used for this purpose. We formulate "natural" requirements on the trend extraction procedure and then show that the inverse F-transform fulfils all of them.
Comments
Technical Report: UTEP-CS-12-46
Published in: M. Inuiguchi, Y. Kusunoki, and M. Seki (eds.), Proceedings of the 15th Czech-Japan Seminar on Data Analysis and Decision Making under Uncertainty CJS'2012, Osaka, Japan, September 24-27, 2012.