Labeled Neuro-Fuzzy Classifier - Université Pierre et Marie Curie Accéder directement au contenu
Article Dans Une Revue Asian Journal of Information Technology Année : 2005

Labeled Neuro-Fuzzy Classifier

Mohamed Nemissi
  • Fonction : Auteur
Hamid Seridi
  • Fonction : Auteur
Herman Akdag
  • Fonction : Auteur
  • PersonId : 903641

Résumé

This study presents a model of Neuro-Fuzzy classification, which its conception is inspired from the labeled classification using Neural Networks. This last aims to improve the classification performances and to accelerate the training of the used classifier. It is based on the addition of a set of labels to all training examples. Tests will be then carried out with each of these labels to classify a new example. The advantage of this approach is the simplicity of its implementation, which does not require modification of the training algorithm. The proposed model is based on the use of this method with the NFC (Neuro Fuzzy Classifier). To appreciate its performances, tests are carried out on the Iris and human tight data basis by the NFC with and withwout labels.
Fichier non déposé

Dates et versions

hal-01172272 , version 1 (07-07-2015)

Identifiants

  • HAL Id : hal-01172272 , version 1

Citer

Mohamed Nemissi, Hamid Seridi, Herman Akdag. Labeled Neuro-Fuzzy Classifier. Asian Journal of Information Technology, 2005, 4 (9), pp.868-872. ⟨hal-01172272⟩
59 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More