Previous |  Up |  Next


optimization; censored data; product-limit estimate; consistency
In the framework of a stochastic optimization problem, it is assumed that the stochastic characteristics of optimized system are estimated from randomly right-censored data. Such a case is frequently encountered in time-to-event or lifetime studies. The analysis of precision of such a solution is based on corresponding theoretical properties of estimated stochastic characteristics. The main concern is to show consistency of optimal solution even in the random censoring case. Behavior of solutions for finite data sizes is studied with the aid of randomly generated example.
[1] Breslow, N., Crowley, J. E.: A large sample study of the life table and product limit estimates under random censorship. Ann. Statist. 2 (1974), 437-453. DOI 10.1214/aos/1176342705 | MR 0458674 | Zbl 0283.62023
[2] Dupačová, J., Wets, R.: Asymptotic behaviour of statistical estimates and optimal solutions of stochastic optimization problems. Ann. Statist. 16 (1984), 1517-1549. DOI 10.1214/aos/1176351052 | MR 0964937
[3] Houda, M., Kaňková, V.: Empirical estimates in economic and financial optimization problems. Bull. Czech Econometr. Soc. 19 (2012), 50-69.
[4] Kalbfleisch, J. D., Prentice, R. L.: The Statistical Analysis of Failure Time Data. Second edition. Wiley, New York 2002. MR 1924807
[5] Kaňková, V.: Empirical estimates in stochastic optimization via distribution tails. Kybernetika 46 (2010), 459-471. MR 2676083 | Zbl 1225.90092
[6] Rejto, L.: On fixed censoring model and consequences for the stochastic case. In: Trans. 9th Prague Conference on Stochastic Decision Functions 1982, Academia, Prague 1983, pp. 141-147. MR 0757919
[7] Robbins, H., Siegmund, D.: Boundary crossing probabilities for a Wiener process and sample sums. Ann. Math. Statist. 41 (1970), 1410-1429. DOI 10.1214/aoms/1177696787 | MR 0277059
[8] Volf, P.: On precision of optimization in the case of incomplete information. Bull. Czech Econometr. Soc. 19 (2012), 170-184.
Partner of
EuDML logo