ویراساینس.
سند علمی

Nearest prototype classifier designs: An experimental study

James C. Bezdek, Ludmila I. Kuncheva

چکیده

We compare eleven methods for finding prototypes upon which to base the nearest Ž prototype classifier. Four methods for prototype selection are discussed: Wilson Hart a . condensation error-editing method , and three types of combinatorial search random search, genetic algorithm, and tabu search. Seven methods for prototype extraction are discussed: unsupervised vector quantization, supervised learning vector quantization Ž . with and without training counters , decision surface mapping, a fuzzy version of vector quantization, c-means clustering, and bootstrap editing. These eleven methods can be usefully divided two other ways: by whether they employ preor postsupervision; and by whether the number of prototypes found is user-defined or ‘‘automatic.’’ Generalization error rates of the 11 methods are estimated on two synthetic and two real data sets. Offering the usual disclaimer that these are just a limited set of experiments, we feel confident in asserting that presupervised, extraction methods offer a better chance for success to the casual user than postsupervised, selection schemes. Finally, our calculations do not suggest that methods which find the ‘‘best’’ number of prototypes ‘‘automatically’’ are superior to methods for which the user simply specifies the number of prototypes. 2001 John Wiley & Sons, Inc.

اسناد مشابه

  • Nearest-prototype Classifier Design by Deterministic Annealing with Random Class Labels∗

    سند علمی۱۹۹۹Ertem Tuncel, Kenneth Rose

    The design of nearest-prototype (NP) classifiers is a challenging problem because of the prevalence of poor local minima, and the piecewise constant nature of the cost function which is incompatible with gradientbased techniques. This paper extends the deterministic annealing (DA

  • Nearest prototype classifier designs: An experimental study

    مقاله‌ی دسترسی آزاد۲۰۰۱International Journal of Intelligent SystemsJames C.Bezdek, Ludmila I.Kuncheva

  • Evolutionary Design of Nearest Prototype Classifiers

    سند علمی۲۰۰۴J. HeuristicsFernando Fernández, Pedro Isasi Viñuela

    In pattern classification problems, many works have been carried out with the aim of designing good classifiers from different perspectives. These works achieve very good results in many domains. However, in general they are very dependent on some crucial parameters involved in t

  • Fuzzy Nearest Prototype Classifier Applied to Speaker Identification

    سند علمی۱۹۹۹Dat Tran, Michael Wagner, Tongtao Zheng

    In a vector quantisation (VQ) based speaker identification system, a speaker model is created for each speaker from the training speech data by using the k-means clustering algorithm. For an unknown utterance analysed into a sequence of vectors, the nearest prototype classifier i

  • Differential evolution based nearest prototype classifier with optimized distance measures for the features in the data sets

    سند علمی۲۰۱۳Expert Syst. Appl.David Koloseni, Jouni Lampinen, Pasi Luukka

    In this paper a further generalization of differential evolution based data classification method is proposed, demonstrated and initially evaluated. The differential evolution classifier is a nearest prototype vector based classifier that applies a global optimization algorithm,

  • Some Notes on Twenty One (21) Nearest Prototype Classifiers

    سند علمی۲۰۰۰SSPR/SPRJames C. Bezdek, Ludmila I. Kuncheva

    Comparisons made in two studies of 21 methods for finding prototypes upon which to base the nearest prototype classifier are discussed. The criteria used to compare the methods are by whether they: (i) select or extract point prototypes; (ii) employ preor post-supervision; and (i