scientific article; zbMATH DE number 1178219
The aim of the book is to investigate a generalized theory of probability, viz. the theory of interval probability, where each event \(A\) is described by an interval \([L(A);U(A)]\) instead of a single real number \(p(A)\). The author focused his research interest on the existence and construction of optimal test algorithms for the general interval probability calculus developed by \textit{K. Weichselberger} [see V. Mammitzsch et al. (eds.), Symposia Gaussiana, 47-64 (1995; Zbl 0843.60001); Lect. Notes Stat. 109, 391-409 (1996; Zbl 0840.60003)], called \(F\)-probability. The level-\(\alpha\) maximin criterion, relying on the upper limit of the interval valued probabilities of errors of the first and second type, is mainly used to establish the connections with the ``classical statistics based on classical probabilities. The results obtained in this framework can be interpreted as a discussion of the existence and development of level-\(\alpha\) maximin tests in the case of complex, nonparametric, composite hypotheses. For testing with interval probability in robust statistics, there is considered the special case of (in some sense continuous) \(C\)-probabilities. \textit{P. J. Huber} and \textit{V. Strassen} [see Ann. of Statist. 1, 251-263 (1973; Zbl 0259.62008)] showed that there exist two classical probabilities (called globally least favourable pairs) which represent the whole testing problem. The so-called ``necessity theorem of Huber and Strassen led to the conjecture that an efficient theory of testing cannot be based on more general classes of interval probabilities. An important result comprised within the present book, related to this conjecture, is that the condition of \(C\)-probabilities is by no means necessary for the development of a powerful theory of hypotheses testing based on interval probabilities, and that it is possible to consider for this just \(F\)-probability fields. Another major topic of research in this book is a detailed study of the general multivariate case on finite sample-spaces. Finally, the contents of the book are organized in six chapters, with the following titles: 1. Interval probabilities; 2. Basic notions on the testing theory for interval probabilities; 3. Level-\(\alpha\) maximin tests for \(F\)-probabilities and least favourable pairs: general existence theorems; 4. Level-\(\alpha\) maximin tests for \(F\)-probabilities with finite sample-space; 5. The form of the fundamental lemma of the alternative problem for \(F\)-probabilities with finite sample-space. Consequences; 6. Level-\(\alpha\) maximin tests for \(F\)-probabilities and the model construction of the \(F\)-probability field structures.
- Minimum distance estimation in imprecise probability models
- Tests and optimal guarded weights of evidence
- Expected utility within a generalized concept of probability -- a comprehensive framework for decision making under ambiguity
- The theory of interval-probability as a unifying concept for uncertainty
- Neyman-Pearson testing under interval probability by globally least favorable pairs: Reviewing Huber-Strassen theory and extending it to general interval probability
- Kurt Weichselberger's contribution to imprecise probabilities and statistical inference
- On the efficient application of F-PRI. A contribution to statistics in the frame of a more general concept of probability.
- scientific article; zbMATH DE number 4007456 (Why is no real title available?)
- An optimal curtailment of a randomized ump test for a proprtion
- On optimal tests for general interval-hypotheses
- Optimal Predictive Tests
- Numerical Software with Result Verification
- Optimal Range for the iid Test Based on Integration Across the Correlation Integral
- A minimum distance estimator in an imprecise probability model -- computational aspects and applications
- Thomas Augustin's contributions to imprecise probability and statistics
- Data-based decisions under imprecise probability and least favorable models
- Finite approximations of data-based decision problems under imprecise probabilities
This page was built for publication:
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4398693)
