Title | ||
---|---|---|
Flexible adaptive filtering by minimization of error entropy bound and its application to system identification |
Abstract | ||
---|---|---|
It has been shown that using minimum error entropy as the cost function leads to important performance gains in adaptive filtering, especially when the Gaussianity assumptions on the error distribution do not hold. In this paper, we show that by using the entropy bound rather than the entropy, we can derive an efficient algorithm for supervised training. We demonstrate its effectiveness by a system identification problem using a generalized Gaussian noise model. |
Year | DOI | Venue |
---|---|---|
2010 | 10.1109/ICASSP.2010.5495347 | ICASSP |
Keywords | Field | DocType |
identification,system identification,learning (artificial intelligence),error distribution,error entropy bound,adaptive filters,adaptive filtering,gaussian processes,supervised training,gaussian noise model,flexible adaptive filtering,entropy,minimum error entropy,cost function,adaptive systems,estimation,upper bound,adaptive filter,learning artificial intelligence,parameter estimation,minimization,gaussian distribution | Mathematical optimization,Pattern recognition,Computer science,Binary entropy function,Adaptive filter,Artificial intelligence,Gaussian process,Principle of maximum entropy,Adaptive algorithm,Estimation theory,System identification,Gaussian noise | Conference |
ISSN | ISBN | Citations |
1520-6149 E-ISBN : 978-1-4244-4296-6 | 978-1-4244-4296-6 | 5 |
PageRank | References | Authors |
0.49 | 3 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Xi-Lin Li | 1 | 547 | 34.85 |
Tülay Adali | 2 | 1690 | 126.40 |