University of Borås

Borås Academic Digital Archive (BADA) >
Forskningspublikationer / Research Publications >
Institutionen Handels- och IT-högskolan / School of Business and IT (HIT) >
Informatik / Informatics >
Konferensbidrag / Conference papers (Informatics) >

Please use this identifier to cite or link to this item: http://hdl.handle.net/2320/11277

Files in This Item:

File Description SizeFormat
ijcnn_final_work.pdf196.81 kBAdobe PDFView/Open
Title: Producing Implicit Diversity in ANN Ensembles
Authors: Johansson, Ulf
Löfström, Tuve
Department: University of Borås. School of Business and IT
Issue Date: 10-Jun-2012
Citation: Neural Networks (IJCNN), The 2012 International Joint Conference on
ISSN: 2161-4393
ISBN: 978-1-4673-1488-6
978-1-4673-1489-3
Pages: 1-8
Publisher: IEEE
Media type: text
Publication type: conference paper, peer reviewed
Keywords: Artificial neural networks
implicit diversity
ANN ensemble technique
Computer Science
Subject Category: Subject categories::Engineering and Technology::Computer and Information Science::Computer Science::Software Engineering
Research Group: CSL@BS
Area of Research: Computer Science
Ensemble Learning
Neural Networks
Strategic Research Area: Business and IT
Abstract: Combining several ANNs into ensembles normally results in a very accurate and robust predictive models. Many ANN ensemble techniques are, however, quite complicated and often explicitly optimize some diversity metric. Unfortunately, the lack of solid validation of the explicit algorithms, at least for classification, makes the use of diversity measures as part of an optimization function questionable. The merits of implicit methods, most notably bagging, are on the other hand experimentally established and well-known. This paper evaluates a number of straightforward techniques for introducing implicit diversity in ANN ensembles, including a novel technique producing diversity by using ANNs with different and slightly randomized link structures. The experimental results, comparing altogether 54 setups and two different ensemble sizes on 30 UCI data sets, show that all methods succeeded in producing implicit diversity, but that the effect on ensemble accuracy varied. Still, most setups evaluated did result in more accurate ensembles, compared to the baseline setup, especially for the larger ensemble size. As a matter of fact, several setups even obtained significantly higher ensemble accuracy than bagging. The analysis also identified that diversity was, relatively speaking, more important for the larger ensembles. Looking specifically at the methods used to increase the implicit diversity, setups using the technique that utilizes the randomized link structures generally produced the most accurate ensembles.
DOI: 10.1109/IJCNN.2012.6252713
URI: http://hdl.handle.net/2320/11277
Sustainable development: -
Appears in Collections:Konferensbidrag / Conference papers (Informatics)

SFX Query

All items in Borås Academic Digital Archive are protected by copyright, with all rights reserved.

 

DSpace Software Copyright © 2002-2010  The DSpace Foundation