Speciality:
01.01.09 (Discrete mathematics and mathematical cybernetics)
Birth date:
19.10.1952
E-mail: Keywords: mathematical statistics; minimax procedures; game theory; 2-person games; decision theory, game-theory methods for estimation of knowledge level.
Subject:
The game theoretic approach to setting and solution of statistical problems is found for confidence interval estimation problem of parameter of the binomial and Hyper geometric distributions. The confidence probability and the minimax decision function are constructed for any set of admissible intervals. One half of the consructed confidence intervals is shorter than classical CIs for the same confidence probabilities. The minimax decision function under consideration enables observation number to be reduced by about 1.5 times from the classical approach for the same reliability and accuracy.
Main publications:
Lutsenko M. M., Maloshevskii S. G. Minimax confidence intervals for the binomial parameter // J. Statist. Plann. Inference, 2003, 113(1), 67–77.
Lutsenko M. M. Statistical games for discrete distributions // Lecture Notes in Econom. and Math. Systems, 2002, 510, 533–548.
Lutsenko M. M., Ivanov M. A. Minimax confidence intervals for the parameter of hypergeometric distribution // Automat. Remote Control, 2000, 61(7), Part 1, 1125–1132.
Lutsenko M. M. Game of timing with additional discontinuity lines // Intern. Conference "Logic, Game Theory and Social Choice". St.Petersburg, 2001, 158–163.
Lutsenko M. M. A game-theoretic method for the estimation of a parameter of the binomial law //Theoret. Probab. Appl., 1990, 35, 467–477.
