2017
DOI: 10.1049/el.2017.2364
|View full text |Cite
|
Sign up to set email alerts
|

Low‐complexity cross‐validation design of a linear estimator

Abstract: Linear signal estimators have extensive applications. Under the minimum mean squared error (MMSE) criterion, the linear MMSE (LMMSE) estimator is optimal but requires knowledge of the covariance matrices. The sample matched filter generally performs worse but requires less a priori knowledge. A composite estimator that combines the sample LMMSE estimator and matched filter is studied, which may lead to noticeable improvements in performance. It is shown that such a gain can be achieved by low-complexity parame… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
4
0

Year Published

2018
2018
2021
2021

Publication Types

Select...
2
2

Relationship

0
4

Authors

Journals

citations
Cited by 4 publications
(4 citation statements)
references
References 8 publications
0
4
0
Order By: Relevance
“…For example, [51] derives analytical LOOCV solutions to minimize the MSE of the CM estimation for Gaussian data, [69] proposes a CV solution to optimize the inverse covariance matrix (a.k.a. precision matrix) estimation, while [21], [70] derive LOOCV solutions for the application of minimum mean square error (MMSE) filtering. In this paper, we extend these works to obtain data-driven and low-complexity approaches for choosing ρ for the MVDR beamformer, by finding a suitable CV cost as an estimate of J (Σ) in ( 6) and its fast evaluation.…”
Section: B Covariance (Scatter) Matrix Estimatormentioning
confidence: 99%
“…For example, [51] derives analytical LOOCV solutions to minimize the MSE of the CM estimation for Gaussian data, [69] proposes a CV solution to optimize the inverse covariance matrix (a.k.a. precision matrix) estimation, while [21], [70] derive LOOCV solutions for the application of minimum mean square error (MMSE) filtering. In this paper, we extend these works to obtain data-driven and low-complexity approaches for choosing ρ for the MVDR beamformer, by finding a suitable CV cost as an estimate of J (Σ) in ( 6) and its fast evaluation.…”
Section: B Covariance (Scatter) Matrix Estimatormentioning
confidence: 99%
“…The grid search algorithm will find the best number of k in the range of 1 to 50 with the step of 1. To obtain the value of accuracy itself, a validation mechanism called K-fold Cross-Validation (CV) [15] with 5 fold, 10 fold, 20 fold, and leave one out (LOO) are used. Accuracy of a model can be calculated with the formula that is shown in equation 9 [16].…”
Section: Performance Measurement and Comparisonmentioning
confidence: 99%
“…Eqn. (47) can be exploited to compute the closed-form LOOCV solution quickly. From (47), the most involved computation for finding the solution of the optimization problem (15) can be implemented as…”
Section: Loocv Choice For Ols-based Covariance Estimationmentioning
confidence: 99%
“…Note that in contrast to the cross-validation methods in [35] and [36] which choose shrinkage factors by a grid search for optimizing the signal estimation performance, the method proposed in this paper has an analytical solution and optimizes covariance matrix estimation. It also differs from [47] which targets the design of a signal estimator that shrinks the sample LMMSE filter toward the matched filter. Example 6: Application to MVDR beamforming: Finally, we show an example application to minimum variance distortionless response (MVDR) beamforming [31], [33].…”
Section: Numerical Examplesmentioning
confidence: 99%