Published June 1981 | Version Published
Working Paper Open

Consistency of Generalized M-Estimators

Creators

Abstract

The consistency of M-estimators in a very general setup is proven under weak assumptions. A one-dimensional result using a quasiconvexity assumption is obtained and applied to get a result on consistency of redescending M-estimators. A result valid in higher dimensions is obtained using a law of large numbers for semicontinuous function-valued random variables. This is applied to minimum absolute deviation regression.

Additional Information

I would like to thank Professor Robert L. Taylor for guiding me through Banach space versions of the law of large numbers.

Attached Files

Published - sswp390.pdf

Files

sswp390.pdf

Files (420.0 kB)

Name Size Download all
md5:5328f99f38b2d54b68fca47c71ad210b
420.0 kB Preview Download

Additional details

Identifiers

Eprint ID
82119
Resolver ID
CaltechAUTHORS:20171005-131210701

Dates

Created
2017-10-06
Created from EPrint's datestamp field
Updated
2019-10-03
Created from EPrint's last_modified field

Caltech Custom Metadata

Caltech groups
Social Science Working Papers
Series Name
Social Science Working Paper
Series Volume or Issue Number
390