DSpace
 

Research at Sofia University >
Faculty of Mathematics and Informatics >
Papers >

Please use this identifier to cite or link to this item: http://hdl.handle.net/10506/22

Title: Experiments with Two Approaches for Tracking Drifting Concepts
Authors: Koychev, Ivan
Keywords: Machine Learning
Forgetting Models
Issue Date: 2006
Publisher: “Serdica Journal of Computing” 1, Institute of Mathematics and Informatics - BAS
Citation: Koychev I. (2006) Experiments with Two Approaches for Tracking Drifting Concepts” - “Serdica Journal of Computing” 1, Institute of Mathematics and Informatics - BAS.
Abstract: . This paper addresses the task of learning classifier from stream of labelled data. In this case we can face problem that the underling concepts can changes over time. The paper studies two mechanisms developed for dealing with changing concepts. Both are based on the time window idea. The first one forgets gradual, by assigning to the examples weight that gradually decreases over time. The second one uses a statistical test to detect changes in concept and then optimizes the size of time window, aiming to maximise the classification accuracy on the new examples. Both methods are general in nature and can be used with any learning algorithm. The objectives of the conducted experiments were to compare the mechanisms and explore whether they can combined to achieve a synergetic effect. Results from experiments with three basic learning algorithms (kNN, ID3 and NBC) using four datasets are reported and discussed.
URI: http://hdl.handle.net/10506/22
Appears in Collections:Papers

Files in This Item:

File Description SizeFormat
Serdica_draft2006.pdf131.06 kBAdobe PDFView/Open
View Statistics

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

 

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