Dynamic Classifier Selection by Adaptive k-Nearest-Neighbourhood Rule

Lecture notes in computer sciencePublished 1 January 2004

Luca Didaci, Giorgio Giacinto

DOI10.1007/978-3-540-25966-4_17

Source URL

Citations28

SJR quartileQ2

SJR score0.35

SNIP0.55

TL;DR

The use of neighbourhoods of adaptive shape and size are investigated to better cope with the difficulties of a reliable estimation of local accuracies and show that performance improvements can be achieved by suitably tuning some additional parameters.

Abstract

Despite the good results provided by Dynamic Classifier Selection (DCS) mechanisms based on local accuracy in a large number of applications, the performances are still capable of improvement. As the selection is performed by computing the accuracy of each classifier in a neighbourhood of the test pattern, performances depend on the shape and size of such a neighbourhood, as well as the local density of the patterns. In this paper, we investigated the use of neighbourhoods; of adaptive shape and size to better cope with the difficulties of a reliable estimation of local accuracies. Reported results show that performance improvements can be achieved by suitably tuning some additional parameters

Keywords

Computer Science

IEEE Transactions on Pattern Analysis and Machine IntelligenceOn combining classifiers

5,227 Citations1998Josef Kittler, M. Hatef +2 more

IEEE Transactions on Systems Man and CyberneticsMethods of combining multiple classifiers and their applications to handwriting recognition

2,171 Citations1992Longchang Xu, Adam Krzyżak +1 more

On applying these methods to combine several classifiers for recognizing totally unconstrained handwritten numerals, the experimental results show that the performance of individual classifiers can be improved significantly.

IEEE Transactions on Pattern Analysis and Machine IntelligenceDecision combination in multiple classifier systems

1,511 Citations1994Tin Kam Ho, J.J. Hull +1 more

This work proposes three methods based on the highest rank, the Borda count, and logistic regression for class set reranking that have been tested in applications of degraded machine-printed characters and works from large lexicons, resulting in substantial improvement in overall correctness.

Encyclopedia of Machine Learning and Data MiningMultiple Classifier Systems

1,041 Citations2017Claude Sammut, Geoffrey I. Webb

IEEE Transactions on Pattern Analysis and Machine IntelligenceCombination of multiple classifiers using local accuracy estimates

949 Citations1997K. Woods, W. Philip Kegelmeyer +1 more

IEEE Transactions on Systems Man and Cybernetics Part B (Cybernetics)Switching between selection and fusion in combining classifiers: an experiment

351 Citations2002Ludmila I. Kuncheva

A combination of classifier selection and fusion by using statistical inference to switch between the two by offering a discussion on when to combine classifiers and how classifiers selection (static or dynamic) can be misled by the differences in the classifier team.

Image Analysis and Processing

140 Citations1986Virginio Cantoni, S. Levialdi +1 more

The first € price and the £ and $ price are net prices, subject to local VAT, and the €(D) includes 7% for Germany, the€(A) includes 10% for Austria.

IEEE Transactions on Geoscience and Remote SensingMultiple classifier systems for supervised remote sensing image classification based on dynamic classifier selection

115 Citations2002P.C. Smits

Experiments show that the proposed method outperform MCS strategies based on belief functions and the DCS-LA in terms of minimum and maximum class accuracies and kappa coefficient of agreement and is a valid alternative to majority voting.

Lecture notes in computer scienceDynamic Classifier Selection

60 Citations2000Giorgio Giacinto, Fabio Roli

This paper is aimed to provide a theoretical framework for dynamic classifier selection and to define the assumptions under which it can be expected to improve the accuracy of the individual classifiers.

Lecture notes in computer scienceAdaptive selection of image classifiers

52 Citations1997Giorgio Giacinto, Fabio Roli

It is pointed out that adaptive selection does not require the assumption of uncorrelated errors, thus simplifying the choice of classifiers forming a Multiple Classifier System.