Consistency of local linear regression estimator for mixtures with varying concentrations
Autor: | Daniel Horbunov, Rostyslav Maiboroda |
---|---|
Jazyk: | angličtina |
Rok vydání: | 2024 |
Předmět: | |
Zdroj: | Modern Stochastics: Theory and Applications, Vol 11, Iss 3, Pp 359-372 (2024) |
Druh dokumentu: | article |
ISSN: | 2351-6046 2351-6054 |
DOI: | 10.15559/24-VMSTA250 |
Popis: | Finite mixtures with different regression models for different mixture components naturally arise in statistical analysis of biological and sociological data. In this paper a model of mixtures with varying concentrations is considered in which the mixing probabilities are different for different observations. A modified local linear estimation (mLLE) technique is developed to estimate the regression functions of the mixture component nonparametrically. Consistency of the mLLE is demonstrated. Performance of mLLE and a modified Nadaraya–Watson estimator (mNWE) is assessed via simulations. The results confirm that the mLLE technique overcomes the boundary effect typical to the NWE. |
Databáze: | Directory of Open Access Journals |
Externí odkaz: |