Deriving Pseudo-Probabilities of Correctness Given Scores (DPPS)

Jaeger

Studies in Computational Intelligence

et al. 2008

Self Cite

160

Classifier combination methods have proved to be an effective tool to increase the performance of pattern recognition applications. In this chapter we review and categorize major advancements in this field. Despite a significant number of publications describing successful classifier combination implementations, the theoretical basis is still missing and achieved improvements are inconsistent. By introducing different categories of classifier combinations in this review we attempt to put forward more specific directions for future theoretical research. We also introduce a retraining effect and effects of locality based training as important properties of classifier combinations. Such effects have significant influence on the performance of combinations, and their study is necessary for complete theoretical understanding of combination algorithms.

show abstract

Section: The Retraining Effectmentioning

confidence: 99%

Review of Classifier Combination Methods

Jaeger

Studies in Computational Intelligence

et al. 2008

Self Cite

160

show abstract

“…In the applications we have considered in this paper, the matching scores re°ect distance measures between the biometric templates or between the handwritten word image and a lexicon word. These distances can be converted to probabilities, 6,9,14 but this conversion is nontrivial and prone to errors. 3 Snelick et al 20 investigated the combination of three¯ngerprint and one face biometric matchers.…”

Section: Previous Workmentioning

confidence: 99%

On the Difference Between Optimal Combination Functions for Verification and Identification Systems

Int. J. Patt. Recogn. Artif. Intell.

2010

Self Cite

We have investigated di®erent scenarios of combining pattern matchers. The combination problem can be viewed as a construction of a postprocessing classi¯er operating on the matching scores of the combined matchers. The optimal combination algorithm for veri¯cation systems corresponds to the likelihood ratio combination function. It can be implemented by the direct reconstruction of this function with genuine and impostor score density approximations. However, the optimal combination algorithm for identi¯cation systems is difcult to express analytically. We will show that this di±culty is caused by the dependencies between matching scores assigned to di®erent classes by the same classi¯er. The experiments on the large sets of scores from handwritten word recognizers operating on postal images and biometric matchers (NIST biometric score set BSSR1) con¯rm the existence of such dependencies and that the optimal combination functions for veri¯cation and identi¯cation systems are di®erent.

show abstract

Matching Score Fusion Methods

Handbook of Statistics - Machine Learning: Theory and Applications

2013