| Titre : | Learning machines : foundations of trainable pattern-classifying systems |
| Auteurs : | Nils J. Nilsson, Auteur |
| Type de document : | texte imprimé |
| Editeur : | New York : McGraw-Hill, 1965 |
| Collection : | McGraw-Hill series in systems science |
| Format : | XI,137 p. / ill. / 23 cm |
| Note générale : | Bibliogr. at the end of chapters. - Index |
| Langues : | Anglais |
| Index. décimale : | 004.8 (Intelligence artificielle) |
| Tags : | Intelligence artificielle Artificial intelligence |
| Résumé : |
This monograph presents some of the results of research in the new and exciting field of learning machines. A learning machine, broadly defined, is any device whose actions are influenced by past experiences. The present work deals specifically with the theory of a subclass of learning machines, those which can be trained to recognize patterns. Some well-known examples of trainable pattern-classifying systems are the perceptron and the madaline and minos networks. |
| Note de contenu : |
Summary :
1. Trainable pattern classifiers. 2. Some important discriminant functions : their properties and their implementations. 3. Parametric training methods. 4. Some nonparametric training methods for machines. 5. Training theorems. 6. Layered machines. 7. Piecewise linear machines. |
Exemplaires (1)
| Cote | Support | Localisation | Section | Disponibilité | Etat_Exemplaire |
|---|---|---|---|---|---|
| 004.8 NIL | Papier | Bibliothèque Centrale | Informatique | Disponible | Consultation sur place |
