|
On a new estimation method of the Bernoulli regression function
P. Babilua, È. A. Nadaraya Tbilisi Ivane Javakhishvili State University
Abstract:
A new estimator for a Bernoulli regression function based on Bernstein
polynomials is constructed. Its consistency and asymptotic normality are
studied. A criterion for testing the hypothesis on the form of a Bernoulli
regression function and a criterion for testing the hypothesis on the
equality of two Bernoulli regression functions are constructed. The
consistency of these two criteria is studied.
Keywords:
Bernstein polynomial, Bernoulli regression function, consistency, power of a criterion, one-sided alternatives.
Received: 04.09.2020 Accepted: 08.07.2021
Published: 22.04.2022
Citation:
P. Babilua, È. A. Nadaraya, “On a new estimation method of the Bernoulli regression function”, Teor. Veroyatnost. i Primenen., 67:2 (2022), 209–222; Theory Probab. Appl., 67:2 (2022), 163–174
Linking options:
https://www.mathnet.ru/eng/tvp5434https://doi.org/10.4213/tvp5434 https://www.mathnet.ru/eng/tvp/v67/i2/p209
|
|