Multivariate Shannon's entropy for adaptive IIR filtering via kernel density estimators. Issue 15 (1st July 2019)

• Record Type: Journal Article
• Authors: Fantinato, D.G.; Silva, D.G.; Attux, R.; Neves, A.
• Is Part Of: Electronics letters. Volume 55:Issue 15(2019)
• Page Start: 859
• Page End: 861
• Publication Date: 2019-07-01
• Subjects: adaptive filters -- equalisers -- filtering theory -- learning (artificial intelligence) -- optimisation -- mean square error methods -- entropy -- IIR filters -- gradient methods -- higher order statistics; Shannon's entropy -- adaptive IIR -- kernel density estimators -- supervised infinite impulse response adaptive filtering -- approximate gradient‐based approaches -- usual option -- optimisation methods -- mean squared error criterion -- MSE -- noisy scenarios -- Shannon's error entropy -- information theoretic learning criterion -- higher order statistics -- underlying signals -- multivariate entropy definition -- Recursive Prediction Error‐based algorithm -- supervised channel equalisation problem; Electronics -- Periodicals; 621.381
• DOI: 10.1049/el.2019.1430
• Languages: English
• ISSNs: 0013-5194

Abstract

Abstract : In supervised infinite impulse response adaptive filtering, approximate gradient‐based approaches are the usual option among optimisation methods. When based on the mean squared error (MSE) criterion, however, these approaches may present biased solutions in noisy scenarios. In that sense, instead of the MSE, the authors propose the use of Shannon's error entropy, an information theoretic learning criterion, which is able to extract higher order statistics from the underlying signals. In particular, a multivariate entropy definition is considered, which is applied to derive a Recursive Prediction Error‐based algorithm. The performance analyses are carried out in the context of the supervised channel equalisation problem, with results very favourable to the proposal, in high and low noise level environments.