In estimating the parameter θ from a parametrized signal problem (with 0⩽θ⩽L) observed through Gaussian white noise, four useful and computable lower bounds for the Bayes risk are developed. For problems with different L and different signal to noise ratios, some bounds are superior to others. The lower bound obtained from taking the maximum of the four, serves not only as a good lower bound for the Bayes risk but also as a good lower bound for the minimax risks. Threshold behavior of the Bayes risk is also evident, as is shown in the lower bound
Published in:
Information Theory, IEEE Transactions on
(Volume:39
,
Issue:
4
)
Date of Publication:
Jul 1993
- Page(s):
-
1386
-
1394
- ISSN :
-
0018-9448
- INSPEC Accession Number:
-
4592687
- Digital Object Identifier :
-
10.1109/18.243453
- Date of Current Version :
-
06 August 2002
- Issue Date :
-
Jul 1993
- Sponsored by :
-
IEEE Information Theory Society