Combining multiple classifiers by averaging or by multiplying?
By David Martinus Johannes Tax, M. V. Breukelen, Robert P. W. Duin, Josef Kittler. Pattern Recognition, 2000
427 citations recorded by the source corpus.
View Paper: https://www.semanticscholar.org/paper/cfc44747b1591c1bb229ea833742787ae3a8709e
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
- Battista Biggio
- Giorgio Fumera
- Fabio Roli
- Anil K. Jain
- Arun A. Ross
- Tien-Ping Tan
- Yunhong Wang
- Robi Polikar
Citations in the corpus
- Personal Authentication Using Finger Knuckle Surface (2009, background): “The SUM rule has been shown [ 27 ], [28] to be useful for combining matching scores from the correlated feature spaces in which case the errors from the classifiers are independent.”
- Multimodal biometrics: An overview (2004, background): “Fusion at the match score level has been well studied in the literature [17, 18].”
- Toward detecting emotions in spoken dialogs (2005, background): “2) because it achieves good performance in spite of its simplicity and less support by probabilistic interpretation [24], [25].”
- A theoretical and experimental analysis of linear combiners for multiple classifier systems (2005, background): “Of the various combining rules proposed in the literature, linear combiners are the most frequently used [22], [12], [2], [30], [16], [31], [32], [28], [20], [19].”
- Ensemble based systems in decision making (2006, background): “Does the no-free-lunch theorem hold for combination rules as well? Many studies have been conducted to answer this question, including [18], [27], [28], [31], [36], [37], [42], [71], [72], among others.”
- Products of Random Latent Variable Grammars (2010, background): “A great deal has been written on the topic of products versus sums of probability distributions for joint prediction (Genest and Zidek, 1986; Tax et al., 2000).”
- Combining Face and Iris Biometrics for Identity Verification (2003, background): “It has been shown that the simple sum rule gives very good accuracy in combining multiple biometric systems [5][6][7][19].”