Consistency of logistic regression coefficient estimates calculated from a training sample
The paper deals with logistic regression when the sample is split into training and holdout sub-samples. Under the assumption that, asymptotically, the ratio of the sizes of the two sub-samples is approximately fixed, we prove that the logistic regression coefficient MLEs calculated from the training sample are consistent.
Year of publication: |
2003
|
---|---|
Authors: | Hadjicostas, Petros |
Published in: |
Statistics & Probability Letters. - Elsevier, ISSN 0167-7152. - Vol. 62.2003, 3, p. 293-303
|
Publisher: |
Elsevier |
Keywords: | Haberman's conditions Maximum likelihood estimates Sample splitting |
Saved in:
Saved in favorites
Similar items by person
-
Improper and proper posteriors with improper priors in a Poisson-gamma hierarchical model
Hadjicostas, Petros, (1999)
-
Maximizing proportions of correct classifications in binary logistic regression
Hadjicostas, Petros, (2006)
-
One-sided elasticities and technical efficiency in multi-output production: A theoretical framework
Hadjicostas, Petros, (2006)
- More ...