The University of Dublin | Trinity College -- Ollscoil Átha Cliath | Coláiste na Tríonóide
Trinity's Access to Research Archive
Home :: Log In :: Submit :: Alerts ::

School of Computer Science and Statistics >
Computer Science >
Computer Science Technical Reports >

Please use this identifier to cite or link to this item:

Title: Blame-Based Noise Reduction: An Alternative Perspective on Noise Reduction for Lazy Learning
Author: Pasquier, François-Xavier
Delany, Sarah Jane
Cunningham, Pádraig
Sponsor: Science Foundation Ireland
Enterprise Ireland
Keywords: Nearest-Neighbour Classifiers
Noise Reduction
Issue Date: 22-Feb-2005
Publisher: Trinity College Dublin, Department of Computer Science
Citation: Pasquier, François-Xavier; Delany, Sarah Jane; Cunningham, Pádraig. 'Blame-Based Noise Reduction: An Alternative Perspective on Noise Reduction for Lazy Learning'. - Dublin, Trinity College Dublin, Department of Computer Science, TCD-CS-2005-29, 2005, pp17
Series/Report no.: Computer Science Technical Report
Abstract: In this paper we present a new perspective on noise reduction for nearest-neighbour classifiers. Classic noise reduction algorithms such as Repeated Edited Nearest Neighbour remove cases from the training set if they are misclassified by their nearest neighbours in a leave-one-out cross validation. In the approach presented here, cases are identified for deletion based on their propensity to cause misclassifications. This approach was originally identified in a case-based spam filtering application where it became clear that certain training examples were damaging to the accuracy of the system. In this paper we evaluate the general applicability of the approach on a large variety of datasets and show that it generally beats the classic approach. We also compare the two techniques on artificial noise and show that both are far from perfect at removing noise and that there remains scope for further research in this area.
Appears in Collections:Computer Science Technical Reports

Files in This Item:

File Description SizeFormat
TCD-CS-2005-29.pdf161.63 kBAdobe PDFView/Open

This item is protected by original copyright

Please note: There is a known bug in some browsers that causes an error when a user tries to view large pdf file within the browser window. If you receive the message "The file is damaged and could not be repaired", please try one of the solutions linked below based on the browser you are using.

Items in TARA are protected by copyright, with all rights reserved, unless otherwise indicated.


Valid XHTML 1.0! DSpace Software Copyright © 2002-2010  Duraspace - Feedback