Grant detail

Theory of selecting a robust procedure for finite samples of contaminated data (1991-1993)

Principal investigator: prof. RNDr. Jan Ámos Víšek CSc.
Collaborators:
Description: The project has proposed several possibilities how to select from the offer of robust estimating regression model the most appropriate one for the given contaminated data.
Participation:
Work in grant:
Web link:
Finance: GA CSAV, číslo grantu 27557
End date: 12/1993
Publications:

Aguilar, L., Rubio,A., J.A. Víšek: Testing for differences between estimates of models.

Quintana, F., Rubio, A., J. A. Víšek.: Sensitivity analysis of M-Estimators of nonlinear regression model,

Rubio, A., J.A. Víšek: Discriminability of robust test under heavy contamination.

Rubio, A., J.A. Víšek: Efficiency rate and local deficiency of Huber's estimators of location and of alpha-estimators.

Víšek, J.Á: Asymptotic distribution of simple estimate for rejective, Sampford amd successive sampling.

Víšek, J.Á. : A proposal of model selection: Stability of linear regression model.

Víšek, J.Á. : Adaptive estimation in linear regression model. Part I. Consistency.

Víšek, J.Á. : Adaptive estimation in linear regression model. Part II. Asymptotic normality.

Víšek, J.Á. : Adaptive maximum-likelihood-like estimation in linear regression model. Part I. Consistency.

Víšek, J.Á. : Adaptive maximum-likelihood-like estimation in linear regression model. Part II. Asymptotic normality.

Víšek, J.Á. : Stability of regression model estimates with respect to subsamples.

Víšek, J.A.: On the role of Contamination Level and the least Favorable Behavior of Gross-error sensitivity.

Conferences:

Partners

Deloitte
Česká Spořitelna

Sponsors

CRIF
McKinsey
Patria Finance
EY