Abstract
Huber and Strassen have shown in [2] that a composite testing problem can be replaced by an equivalent single one (in terms of least favourable pairs of distributions), when both hypotheses are given by 2-alternating capacities. In this paper a rather general technique is presented in order to construct the least favourable distribution, if one of the hypotheses is a simple one. The mentioned technique applies not only to the total-variation and the ε-contamination model but also to Prochorov-neighbourhoods of distributions on the real line.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Huber, P.J.: A robust version of the probability ratio test, Ann. Math. Stat. 35 (1965), 73–101.
Huber, P.J. and V. Strassen: Minimax tests and Neyman-Pearson lemma for capacities, Ann. Stat. 1 (1973), 251–263 and Ann. Stat. 2 (1974), 223–224.
Rieder, H.: Zur finiten und asymptotischen Theorie robuster Tests, Dissertation, Albert-Ludwigs-Universität Freiburg i. Br. (1976).
Ă–sterreicher, F.: On the construction of least favourable pairs of distributions, submitted to Z. f. Wahrscheinlichkeitstheorie verw. Geb. (1976).
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1978 Springer-Verlag
About this paper
Cite this paper
Österreicher, F. (1978). On the construction of least favourable distributions. In: Kallianpur, G., Kölzow, D. (eds) Measure Theory Applications to Stochastic Analysis. Lecture Notes in Mathematics, vol 695. Springer, Berlin, Heidelberg. https://doi.org/10.1007/BFb0062672
Download citation
DOI: https://doi.org/10.1007/BFb0062672
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-09098-4
Online ISBN: 978-3-540-35556-4
eBook Packages: Springer Book Archive