Newsgroups: comp.ai.neural-nets
Path: cantaloupe.srv.cs.cmu.edu!nntp.club.cc.cmu.edu!godot.cc.duq.edu!hudson.lm.com!news.pop.psu.edu!news.cac.psu.edu!howland.reston.ans.net!pipex!bt!btnet!uknet!festival!ajmy
From: ajmy@festival.ed.ac.uk (A Myles)
Subject: Re: NN Vs Stats......
References: <srjg7930-2601950903020001@ruger-59.slip.uiuc.edu>
Message-ID: <D365w0.JL2@festival.ed.ac.uk>
Organization: Edinburgh University
Date: Sun, 29 Jan 1995 13:38:23 GMT
Lines: 30

srjg7930@uxa.cso.uiuc.edu writes:

>which is geometrically distorted.  I believe that there isn't any
>statistical method to do so.  When one encounters missing data in any
>typical regression analyses (just as an example) the usual procedures are
>(short of tossing out the data or resampling) to fill in the missing data 
>by hand, i.e. by using the sample mean, assuming the missing data have the
>same value as the last know datum point, etc.  

I've not looked at most of these for quite some time now, but a few
good recent articles for regression are:

(Don't know about vision, it's not my thang, but I do know that Ahmad
and Co. have had some success here using the ML methods, covered in
some recent NIPS.)

"Missing Data, Imputation, and the Bootstrap", B. Efron,
JASA 1994, pp. 463-479

"A Bootstrap Method for Using Imputation techniques for Data with
Missing Values", A. L. Bello, Biom. J. 1994, pp. 453-464

"Regression With Missing X's: A Review", R.J. Little,
JASA 1992, pp. 1227-1237

"Statistical Analysis with Missing Data", R.J.A. Little, D.B. Rubin,
Wiley, 1987, ISBN 0-471-80254-9

"Multiple Imputation for Nopnresponse in Sample Surveys", D.B. Rubin,
Wiley, 1987, ISBN 0-471-08705-X
