Identifying the finite dimensionality of curve time series

Abstract

The curve time series framework provides a convenient vehicle to accommodate some nonstationary features into a stationary setup. We propose a new method to identify the dimensionality of curve time series based on the dynamical dependence across different curves. The practical implementation of our method boils down to an eigenanalysis of a finite-dimensional matrix. Furthermore, the determination of the dimensionality is equivalent to the identification of the nonzero eigenvalues of the matrix, which we carry out in terms of some bootstrap tests. Asymptotic properties of the proposed method are investigated. In particular, our estimators for zero-eigenvalues enjoy the fast convergence rate n while the estimators for nonzero eigenvalues converge at the standard $\sqrt{n}$-rate. The proposed methodology is illustrated with both simulated and real data sets.

arXiv commentPublished in at http://dx.doi.org/10.1214/10-AOS819 the Annals of Statistics (http://www.imstat.org/aos/) by the Institute of Mathematical Statistics (http://www.imstat.org)

References in corpus (1)

• Common functional principal components

Cited by in corpus (6)

• Modelling and forecasting daily electricity load curves: a hybrid approach
• Factor modeling for high-dimensional time series: Inference for the number of factors
• An autocovariance-based learning framework for high-dimensional functional time series
• Factor modelling for high-dimensional functional time series
• Finite Sample Theory for High-Dimensional Functional/Scalar Time Series with Applications
• AR-sieve Bootstrap for High-dimensional Time Series