Introduction to statistical pattern recognition (2nd ed.)
Introduction to statistical pattern recognition (2nd ed.)
The SVD and reduced rank signal processing
Signal Processing - Theme issue on singular value decomposition
Nonlinear total variation based noise removal algorithms
Proceedings of the eleventh annual international conference of the Center for Nonlinear Studies on Experimental mathematics : computational issues in nonlinear science: computational issues in nonlinear science
Recovery of blocky images from noisy and blurred data
SIAM Journal on Applied Mathematics
Performance analysis of the DCT-LMS adaptive filtering algorithm
Signal Processing
A Multilinear Singular Value Decomposition
SIAM Journal on Matrix Analysis and Applications
Optimization by Vector Space Methods
Optimization by Vector Space Methods
Constructing fixed rank optimal estimators with method of best recurrent approximations
Journal of Multivariate Analysis
Estimation of Dependences Based on Empirical Data: Springer Series in Statistics (Springer Series in Statistics)
An optimal filter of the second order
IEEE Transactions on Signal Processing
Correlation estimators based on simple nonlinear transformations
IEEE Transactions on Signal Processing
Method of Hybrid Approximations for Modelling of Multidimensional Nonlinear Systems
Multidimensional Systems and Signal Processing
Constructing fixed rank optimal estimators with method of best recurrent approximations
Journal of Multivariate Analysis
Filtering and compression for infinite sets of stochastic signals
Signal Processing
Hi-index | 0.08 |
A new approach to random signal filtering from observed data is proposed. The differences from the known methods are as follows. First, the signal is estimated by a special iterative procedure aimed at improving the accuracy of estimates obtained for the preceding iterative loops. Second, a new best quadratic estimation problem is solved on each iterative loop, providing better estimation accuracy compared with customary linear least-squares methods. The combination of these two new techniques results in a significant improvement in the filtering procedure performance compared with known methods. We call the proposed approach the method of recurrent best estimators of second degree. The efficiency of the method is illustrated by simulations with signals given by digitized images.