Decomposition of additive random fields
The authors study additive random fields \[ Y_d(t)= \sum_{j=1}^dX_j(t_j),\ t=(t_1,\ldots,t_d)\in [0,1]^d, \] where \(X_j(t_j), j=1,\ldots,d,\) are uncorrelated random processes with zero mean and the same continuous covariance function. The case when all eigenvalues of the covariance operator of the marginal processes equal 1 is rather simple for investigation. However, for other cases of eigenvalues these random fields are substantially more difficult to study. In this case, the decomposition of a random field into the sum of its integral and its centered version is often considered. These two parts are orthogonal in \(L_2([0, 1]^d),\) but, in general, are correlated. The authors propose another decomposition of these random fields, which has orthogonal and uncorrelated summands. For large \(d,\) the corresponding summands are close to the integral and to the centered version of the random field and have small relative mean square errors. The obtained results can be useful for the finite-rank approximation problems with a large parametric dimension \(d.\)
- Approximation complexity of additive random fields
- Approximation complexity of sums of random processes
- Approximation of additive random fields based on standard information: average case and probabilistic settings
- Asymptotic analysis of average case approximation complexity of additive random fields
- Average-case analysis of numerical problems
- scientific article; zbMATH DE number 2078184 (Why is no real title available?)
- Mean Square Truncation Error in Series Expansions of Random Functions
- Small ball probabilities for Gaussian random fields and tensor products of compact operators
This page was built for publication: Decomposition of additive random fields
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2227805)
