METHODS FOR PERFORMANCE ANALYSIS OF SEQUENTIAL STATISTICAL TESTS
Abstract
Sequential statistical tests for simple hypotheses on parameters of probability distributions of independent observations, as well as of Markov chains are considered in the article. Methods for analysis of performance characteristics (I and II type error probabilities, conditional expected sample sizes) of sequential statistical tests are constructed both on the basis of the approximations of test statistics and on the basis of absorbing Markov chain theory. The proposed methods allow assessing the performance characteristics of sequential statistical tests not only for the hypothetical model of data, but also under deviations from this model, which can be used for robustness analysis of sequential tests.
Supporting agencies: The research is supported by the State Research Program «Convergence-2020».
About the Author
A. Yu. Kharin
Belarusian State University, Minsk
Belarus
Belarus
Ph. D. (Physics and Mathematics), Associate Professor
4, Nezavisimosti Ave., 220030
References
1. Wald A. Sequential analysis. New York, John Wiley and Sons, 1947. 212 p.
2. Mukhopadhyay N., Datta S., Chattopadhyay S. Applied Sequential Methodologies. New York, Marcel Dekker, 2004. 452 p.
3. Ghosh B., Sen P. K. (eds.) Handbook of Sequential Analysis. New York, Marcel Dekker, 1991. 637 p.
4. Lai T. Sequential analysis: Some classical problems and new challenges. Statistica Sinica, 2001, vol. 11, pp. 303–408.
5. Huber P., Ronchetti E. Robust Statistics. New York, Wiley, 2009. doi.org/10.1002/9780470434697
6. Kharin A. Robustness analysis for Bayesian sequential testing of composite hypotheses under simultaneous distortion of priors and likelihoods. Austrian Journal of Statistics, 2011, vol. 40, pp. 65–73.
7. Kemeny J. G., Snell J. L. Finite Markov Chains. New York, Springer, 1960. 210 p.
8. Kharin A. Y. An approach to performance analysis of the sequential probability ratio test for the simple hypotheses testing. Vestnik Belorusskogo Gosudarstvennogo Universiteta. Seriya fiziko-matematicheskikh nauk = Proceedings of the Belarusian State University. Physics and Mathematics Sciences, 2002, vol. 1, pp. 92–96 (in Russian).
9. Kharin A. Yu. Robustness of Bayesian and sequential statistical decision rules. Minsk, Belarusian State University, 2013. 207 p. (in Russian).
10. Kharin A. Y. Performance and robustness evaluation in sequential hypotheses testing. Communications in Statistics – Theory and Methods, 2016, vol. 45, no. 6, pp. 1663–1709.
Review
Views: 663
Email this article 