NSF-EPSCOR

 

PROJECT

 
Ash Abebe  Nedret Billor
  Asuman Turkmen  Sai Nudurapati

http://stat.auburn.edu/depth

 

 

 

[Home]  [Project Description]  [Activities]  [Papers]   [Contact Us]  
 

 

 

 

 

Project Description
 

Title: Classification based on Data Depth Ordering

Amount:

$119,977.00

Period:

07/01/2006 - 06/30/2008

PI:

Asheber Abebe, (Co-PI) Nedret Billor (Co-PI),

Grant

NSF-DMS-0604726

Program:

Statistics Program in the Division of Mathematical Sciences and Experimental Program to Stimulate Competitive Research (EPSCoR)

Organization:

National Science Foundation

 

Summary

The problem of classifying entities into one of several groups has been one of the main goals of

many scientific investigations. For instance, identifying a tumor or a case of flu as one of the many different possibilities is potentially life-saving and hence is indespensable to physicians. This is an inherently multivariate problem since measurements are made on several aspects (variables) of the entity as an attempt to best capture its place among others. As it is not generally possible to obtain a measure all variables pertaining to an entity we wish to classify or even get perfect measurements of the measured variables, we perform classification in the presence of uncertainty. It is important that this activity is done in a manner that minimizes the misclassification error rate with efficiency and in a way that is robust to outlying cases.

In developing univariate statistical estimation and testing methods, it is well known that the implementation of rank-based nonparametric procedures results in estimators that are usually robust and efficient. Using statistical depth functions as generalization of univariate ranks to the multivariate case, this project proposes methods for performing multivariate statistical analyses, in particular classification, that promise to be robust and efficient.

The project will give what can be regarded as a complete development of the new classification

technique. Issues of asymptotic optimality, robustness, computational algorithms and finite sample performance will be dealt with step by step and methodically throughout the duration of the project. The project also includes implementations of the newly developed classifier using genetic data sets that are publicly available and comparing its performance to existing classifiers.

 

Activities   
 

1. Presentations

  • JSM05, August, 2005, Minneapolis, Poster presentation by Asuman Turkmen and Sai Nudurapati (Presentation: pdf)

  • Blackwell-Tapia Conference November 3-4, 2006 Poster presentation by Asheber Abebe.

  • IISA Joint Statistical Meeting and International Conference Cochin India, January 02-06 2007 Contributed talk by Sai Nudurapati.

  • ENAR/IMS Spring Meeting, in Atlanta, March 11-14, 2007 by Nedret Billor  

  • The 2007 Meeting of the Classification Society of North America, the University of Illinois at Urbana-Champaign in Urbana, Illinois by Nedret Billor.

  • JSM 07,July 29 - August 2, 2007, Salt Lake City by Asuman Turkmen.

 

Papers  
  Billor, N. , Abebe, A., Turkmen, A. and Nudurapati, S. (2005) Classification based on Data Depth , submitted to Journal of Classification.( pdf  )


Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.