Polynomial Regression under a Mixture of Classical and Berkson Errors

by Oleksandr Liubimov (Taras Shevchenko National University of Kyiv, Faculty of Mathematics and Mechanics)

Tuesday 8 Sept 2026, 17:00 → 18:00 Europe/Kiev

https://us06web.zoom.us/j/87484295028?pwd=n8yEBJHE3tAwOoPpYNEZPGfWNV4SxJ.1 (ONLINE)

Polynomial errors-in-variables regression model under a mixture of classical and Berkson errors is studied. We construct strongly consistent estimators for unknown parameters and show that under mild conditions the estimators are asymptotically normal with a nonsingular asymptotic covariance matrix. Moreover, we find two pairs of asymptotically independent estimators. A numerical example is provided to demonstrate that even under the normality assumption there are no more asymptotically independent pairs.