Abstract: The paper considers the method of antirobast estimation (MAO) of parameters of linear regression equation, based on minimization of Chebyshev distance between calculated and actual values of dependent variable. Unlike the least modules method, which essentially ignores emissions in data, MAO, on the contrary, gravitates towards them. It is shown that, according to experimental results, the number of modulo-maximum approximation errors of the equation is not less than the number of parameters plus one.
Index terms: regression equation, antirobast parameter estimation, approximation errors.


Russia, 659305, Altai region, Biysk,
Trofimova Street, 27, room 404B
Tel. + 7-923-162-93-27
(executive secretary -
Golykh Roman Nikolayevich)
e-mail: info@s-sibsb.ru

The certificate