| Juan-Carlos Perez and Enrique Vidal. Constructive design of LVQ and DSM classifiers. In J. Mira, J. Cabestany, and A. Prieto, editors, New Trends in Neural Computation, Lecture Notes in Computer Science No. 686, pages 335--339. Springer, 1993. |
....published; for instance, Johnson and collaborators in [6] use GAs to select the optimal set of descriptors that are fed to a vector quantization algorithm; Fr anti and collaborators in [13] use a genetic algorithm for codebook generation. An incremental methodology proposed by Perez and Vidal [14], seems to offer the best results for this kind of methodology. This method adds or takes codevectors after presentation of the whole training sample, eliminating those that have not been near any vector in the training sample, and adding as new codevector a vector from the training sample that ....
Juan-Carlos Perez and Enrique Vidal. Constructive design of LVQ and DSM classifiers. In J. Mira, J. Cabestany, and A. Prieto, editors, New Trends in Neural Computation, Lecture Notes in Computer Science No. 686, pages 335--339. Springer, 1993.
....(for a review, see [8] which are still local error gradient descent search algorithms on the space of networks or dictionnaries with different size; or on genetic algorithms [9] but in this case they have got a maximum implicit size. An incremental methodology proposed by Perez and Vidal [10], seems to offer the best results for this kind of methodology methodology. This method adds or takes codevectors after presentation of the whole training sample, eliminating those that have not been near any vector in the training sample, and adding as new codevector a vector from the training ....
Juan Carlos P'erez and Enrique Vida. Constructive design of lvq and dsm classifiers. In J. Mira, J. Cabestany, and A. Prieto, editors, New Trends in Neural Computation, pages 334--339, 1993.
....a maximum implicit weight; this maximum length limits the search space, which might be a problem to find the correct codebook, whose number of levels can be higher. helicasas nuevo.tex; 30 07 1997; 9:34; no v. p.3 4 J. J. Merelo et al. An incremental methodology proposed by Perez and Vidal [14], seems to offer the best results for this kind of methodology. This method adds or takes codevectors after presentation of the whole training sample, eliminating those that have not been near any vector in the training sample, and adding as new codevector a vector from the training sample that ....
Juan Carlos P'erez and Enrique Vida. Constructive design of lvq and dsm classifiers. In J. Mira, J. Cabestany, and A. Prieto, editors, New Trends in Neural Computation, pages 334--339, 1993.
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