D.B. Skalak, Combining nearest neighbor classifiers, Dept. of Computer Science, University of Massachusetts, Amherst MA, PhD Thesis, 1997. Home/Search Document Not in Database Summary Related Articles Check |
This paper is cited in the following contexts:
Decision Committee Learning with Dynamic Integration of Classifiers - Tsymbal (2000) (Correct)
.... that use combination of classifiers include the stacking (stacked generalization) architecture [26] the SCANN method that is based on the correspondence analysis and the nearest neighbor procedure [14] combining the minimal nearest neighbor classifiers within the stacked generalization framework [20], and different versions of resampling (as boosting, bagging, and cross validated resampling) that use one learning algorithm to train different classifiers on subsamples of the training set and then voting to combine the classifications of those classifiers [4,18] Two effective classifiers ....
....as which base classifiers should be used, what attributes should be used at the meta level training, and what combining classifier should be used. Different combining algorithms have been used by various researchers, as simple voting in bagging [4] ID3 for combining nearest neighbor classifiers [20], and the nearest neighbor classification in the space of correspondence analysis results (not directly on the predictions) 14] Common solutions of the classification problem are based on the assumption that the entire space of features for a particular domain area consists of null entropy ....
Skalak, D.B.: Combining Nearest Neighbor Classifiers. Ph.D. Thesis, Dept. of Computer Science, University of Massachusetts, Amherst, MA (1997).
Arbiter Meta-Learning with Dynamic Selection of.. - Tsymbal, Puuronen.. (1999) (Correct)
....Meta Learning In chapter 2, we discussed the arbiter meta learning technique, which uses a kind of voting, arbitration rule, to integrate multiple classifiers both in the one level approach and in the arbiter tree approach. The voting technique, however, has several shortcomings (see for example [13]) From our point of view the most important shortcoming is that the voting technique is unable to take into account the local expertise. When a new instance is difficult to classify, then the average classifier will give a wrong prediction, and the majority vote will more probably result in a ....
Skalak, D.B.: Combining Nearest Neighbor Classifiers. Ph.D. Thesis, Dept. of Computer Science, University of Massachusetts, Amherst, MA (1997)
Arbiter Meta-Learning with Dynamic Selection of Multiple.. - Tsymbal, Puuronen (Correct)
....the arbiter meta learning technique. It uses some variation of the voting (called arbitration rule) to integrate multiple classifiers both in one level approach and in the arbiter tree approach. The voting technique, however, has several shortcomings. Some of them, for example, are considered in [7]. The most important shortcoming for us is that the voting technique is unable to take into account local expertise. If a new instance is difficult to classify, then the average classifier will give a wrong prediction. Using a majority vote of classifiers that give more probably wrong predictions ....
D. Skalak. Combining Nearest Neighbor Classifiers. PhD thesis, University of Massachusets, 1997. Available as Dept. of Computer Science Technical Report 96-89.
Diversity in Ensemble Feature Selection - Tsymbal, Pechenizkiy, Cunningham (2003) (Correct)
No context found.
D.B. Skalak, Combining nearest neighbor classifiers, Dept. of Computer Science, University of Massachusetts, Amherst MA, PhD Thesis, 1997.
Local Feature Selection with Dynamic Integration of Classifiers - Puuronen, Tsymbal (2001) (Correct)
No context found.
Skalak, D. B.: Combining nearest neighbor classifiers, Ph.D. Thesis, Department of Computer Science, University of Massachusetts, Amherst, 1997.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC