Publications
A. Boudou and S. Viguier-Pla (2020)
Principal Components Analysis of a Cyclostationary Random Function.
- Classification
- 60G57, 60G10, 60B15, 60H05
- Keywords
- Cyclostationary random function,
Orthogonal projector, Random measure,
Spectral measure, Stationary process,
Unitary operator
Abstract :
Principal Components Analysis is a well-known method for reduction of
dimension in Data Analysis. Considering a cyclostationary random function, we use
appropriate transformations, based on spectral properties, in order to get a stationary
random function, and then to process to a principal components analysis in the
frequency domain. Then, a cyclostationary function is reconstituted as a summary
of the initial cyclostationary function. Applications on simulated data illustrate the
method.