Covering traditional linear pattern recognition and its nonlinear extension via neural networks from an algorithmic approach, this practical "why-and-how" book provides a refreshing contrast to the thoeretical approaches and "pie-in-the-sky" claims of competing books. It explores multiple-layered preceptrons and describes network types such as functional link, radial basis function, learning vector quantanization, and self-organizing, and also discusses recent clustering methods. Suitable for readers with some background in pattern recognition and neural networks, this accesible volume also serves as a useful reference and resource work.

Под заказ:
	OZON.ru - 4331 руб.	Перейти

Рейтинг книги: 4 из 5, 3 голос(-ов).