Multiple Classifier Systems

Lecture notes in computer sciencePublished 1 January 2004

MARCIALIS, GIAN LUCA, Fabio Roli

DOI10.1007/b98227

Source URL

Citations39

SJR quartileQ2

SJR score0.35

SNIP0.55

Abstract

Recent works about perceptron-based fusion of multiple fingerprint matchers showed the effectiveness of such approach in improving the\nperformance of personal i dentity verification systems. However, to the best of our knowledge, no previous work investigated such fusion\n\tapproach when stringent requirements in terms of verification errors\nare given, and the number of available samples for perceptron training is small. Such investigation can allow to understand for which kind\nof applications such fusion rule can be useful. Reported experiments,based on two benchmark data sets, show that perceptron-based fusion\n\tcan be useful for high security fingerprint verification applications,and it is effective in small-sample-size realistic cases.

Keywords

Computer Science

Lecture notes in computer scienceAn Empirical Comparison of Hierarchical vs. Two-Level Approaches to Multiclass Problems

32 Citations2004Suju Rajan, Joydeep Ghosh

The results show that while there is no clear advantage to either technique in terms of classification accuracy, BHCs typically achieve this performance using fewer classifiers, and have the added advantage of automatically generating a hierarchy of classes.

Lecture notes in computer sciencePhysics-Based Decorrelation of Image Data for Decision Level Fusion in Face Verification

24 Citations2004Josef Kittler, Mohammad Taghi Sadeghi

By decorrelating the information content of the respective data channels, this work enhances the diversity of the resulting face verification experts as well as the performance of the multiple classifier system.

Lecture notes in computer scienceNetwork Intrusion Detection by a Multi-stage Classification System

21 Citations2004L.P. Cordella, Alessandro Limongiello +1 more

A serial multi-stage classification system for facing the problem of intrusion detection in computer networks uses suitable criteria for estimating the reliability of the performed classification, so that, in case of uncertainty, information related to a possible attack are only logged for further processing, without raising an alert for the system manager.

Lecture notes in computer scienceExperimental Study on Multiple LDA Classifier Combination for High Dimensional Data Classification

13 Citations2004Xiaogang Wang, Xiaoou Tang

This paper proposes an approach to generate multiple Principalspace LDA and Null Space LDA classifiers by random sampling on the feature vector and training set to improve LDA for high dimensional data classification.

Lecture notes in computer scienceEnsembles of Classifiers Derived from Multiple Prototypes and Their Application to Handwriting Recognition

6 Citations2004Simon Günter, Horst Bunke

A novel algorithm which combines both approaches to create a multiple classifier system which is experimentally evaluated in the context of hidden Markov model (HMM) based handwritten word recognizers and compared to previously introduced methods which also combine both approaches.

Lecture notes in computer scienceSecond Guessing a Commercial’Black Box’ Classifier by an’In House’ Classifier: Serial Classifier Combination in a Speech Recognition Application

3 Citations2004Fuad Rahman, Yuliya Tarnikova +2 more

This paper describes how an ’in house’ classifier can enhance the performance of a commercial ’black box’classifier using the classic serial multiple classifier combination scheme and presents an application in speech recognition.