Rank-Based Multiple Classifier Decision Combination: A Theoretical Study

Saranlı, Afşar; Demirekler, Mübeccel

doi:10.20965/jaciii.2001.p0037

Cited by 3 publications

(2 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Compared to other methods such as Logistic Regression (correlation approach), these methods are simple as they do not require observation of the classifiers [6]. Correlation method is computationally expensive as we need to correlate input image with all images in the training set.…”

Section: Decision Fusionmentioning

confidence: 99%

See 1 more Smart Citation

Decision Fusion Based on Voting Scheme for IR and Visible Face Recognition

Shahbe

Hati

2007

Computer Graphics, Imaging and Visualisation (CGIV 2007)

View full text Add to dashboard Cite

In this paper we present an evaluation study of decision fusion strategies for infrared (IR) and visible face recognition. Several decision fusion methods based on voting scheme (minimization, product and averaging) are discussed and experiments for various conditions of probe set are performed on two databases with paired IR and visible face imageries. The Eigenfaces and Fisherfaces classification techniques are used to extract the face features and the performance of fusion methods are discussed.

show abstract

Section: Decision Fusionmentioning

confidence: 99%

“…Correlation method is computationally expensive as we need to correlate input image with all images in the training set. Minimizing method for multiple classifiers combination is discussed in [3,6]. For this method, we make use the strength of both the classifiers where the classifier that performs better is considered for recognition.…”

Section: Decision Fusionmentioning

confidence: 99%

Decision Fusion Based on Voting Scheme for IR and Visible Face Recognition

Shahbe

Hati

2007

Computer Graphics, Imaging and Visualisation (CGIV 2007)

View full text Add to dashboard Cite

show abstract

A statistical unified framework for rank-based multiple classifier decision combination

Saranlı

Demirekler

2001

Pattern Recognition

View full text Add to dashboard Cite

On output independence and complementariness in rank-based multiple classifier decision systems

Saranlı

Demirekler

2001

Pattern Recognition

View full text Add to dashboard Cite

This study presents a theoretical analysis of output independence and complementariness between classi ers in a rank-based multiple classi er decision system in the context of the Partitioned Observation Space theory. T o enable such an analysis, an Information Theoretic interpretation of a rank-based multiple classi er system is developed and basic concepts from Information Theory are applied to develop measures for output independence and complementariness. It is shown that output independence of classi ers is not a requirement for achieving complementariness between these classi ers. Namely, output independence does not imply a performance improvement b y combining multiple classi ers. A condition called Dominance is shown to be important instead. The information theoretic measures proposed for output independence and complementariness are justi ed by simulated examples.

show abstract

Rank-Based Multiple Classifier Decision Combination: A Theoretical Study

Abstract: This study presents a theoretical investigation of the rankbased multiple classifier decision problem for closed-set pattern identification.

Cited by 3 publications

References 26 publications

Decision Fusion Based on Voting Scheme for IR and Visible Face Recognition

Decision Fusion Based on Voting Scheme for IR and Visible Face Recognition

A statistical unified framework for rank-based multiple classifier decision combination

On output independence and complementariness in rank-based multiple classifier decision systems

Contact Info

Product

Resources

About