TY - JOUR
T1 - Mixed group ranks
T2 - Preference and confidence in classifier combination
AU - Melnik, Ofer
AU - Vardi, Yehuda
AU - Zhang, Cun Hui
N1 - Funding Information:
This research is partially supported by the US Office of Naval Research, US National Science Foundation and DIMACS. The authors would like to thank Jonathon Phillips for useful conversations on face recognition methodologies and the anonymous reviewers for their helpful comments.
PY - 2004/8
Y1 - 2004/8
N2 - Classifier combination holds the potential of improving performance by combining the results of multiple classifers. For domains with very large numbers of classes, such as biometrics, we present an axiomatic framework of desirable mathematical properties for combination functions of rank-based classifiers. This framework represents a continuum of combination rules, including the Borda Count, Logistic Regression, and Highest Rank combination methods as extreme cases [11], [23], [4], [13]. Intuitively, this framework captures how the two complementary concepts of general preference for specific classifiers and the confidence it has in any specific result (as indicated by ranks) can be balanced while maintaining consistent rank interpretation. Mixed Group Ranks (MGR) is a new combination function that balances preference and confidence by generalizing these other functions. We demonstrate that MGR is an effective combination approach by performing multiple experiments on data sets with large numbers of classes and classifiers from the FERET face recognition study.
AB - Classifier combination holds the potential of improving performance by combining the results of multiple classifers. For domains with very large numbers of classes, such as biometrics, we present an axiomatic framework of desirable mathematical properties for combination functions of rank-based classifiers. This framework represents a continuum of combination rules, including the Borda Count, Logistic Regression, and Highest Rank combination methods as extreme cases [11], [23], [4], [13]. Intuitively, this framework captures how the two complementary concepts of general preference for specific classifiers and the confidence it has in any specific result (as indicated by ranks) can be balanced while maintaining consistent rank interpretation. Mixed Group Ranks (MGR) is a new combination function that balances preference and confidence by generalizing these other functions. We demonstrate that MGR is an effective combination approach by performing multiple experiments on data sets with large numbers of classes and classifiers from the FERET face recognition study.
UR - https://www.scopus.com/pages/publications/3242670824
UR - https://www.scopus.com/pages/publications/3242670824#tab=citedBy
U2 - 10.1109/TPAMI.2004.48
DO - 10.1109/TPAMI.2004.48
M3 - Article
C2 - 15641728
AN - SCOPUS:3242670824
SN - 0162-8828
VL - 26
SP - 973
EP - 981
JO - IEEE Transactions on Pattern Analysis and Machine Intelligence
JF - IEEE Transactions on Pattern Analysis and Machine Intelligence
IS - 8
ER -