Weighted moving averaging revisited: an algebraic approach

An algebraic approach for the selection of weight coefficients for weighted moving averaging is proposed in this paper. The algebraic complexity of the sequence transformed by weighted moving averaging is set as a target criterion for the optimization problem of weight coefficients. A special comput...

Full description

Saved in:
Bibliographic Details
Published in:Computational & applied mathematics 2017-12, Vol.36 (4), p.1545-1558
Main Authors: Landauskas, Mantas, Navickas, Zenonas, Vainoras, Alfonsas, Ragulskis, Minvydas
Format: Article
Language:English
Subjects:
Algebra
Applications of Mathematics
Applied physics
Computation
Computational mathematics
Computational Mathematics and Numerical Analysis
Logic programming
Mathematical Applications in Computer Science
Mathematical Applications in the Physical Sciences
Mathematics
Mathematics and Statistics
Time series
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:An algebraic approach for the selection of weight coefficients for weighted moving averaging is proposed in this paper. The algebraic complexity of the sequence transformed by weighted moving averaging is set as a target criterion for the optimization problem of weight coefficients. A special computational setup is constructed in order to tackle the inevitable additive noise for real-world time series. Computational experiments prove that the proposed approach can outperform time series predictors based on classical moving averaging.
ISSN:0101-8205
2238-3603
1807-0302
DOI:10.1007/s40314-016-0309-9