We now begin two chapters presenting network architectures that differ from the more traditional networks of ► Chaps. 15 and 18 . We begin, ► Sects. 19.1 and 19.2, with two examples of Kohonen networks, first the winner-takes-all network and then a network that learns prototypes. The Grossberg outstar architecture and supervised Hebbian learning are demonstrated in ► Sects. 19.3 and 19.4 respectively. We conclude the chapter, ► Sect. 19.5, with an associative memory network.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Alternative Network Architectures: Prototypes and Classifiers

  • George F. Luger

摘要

We now begin two chapters presenting network architectures that differ from the more traditional networks of ► Chaps. 15 and 18 . We begin, ► Sects. 19.1 and 19.2, with two examples of Kohonen networks, first the winner-takes-all network and then a network that learns prototypes. The Grossberg outstar architecture and supervised Hebbian learning are demonstrated in ► Sects. 19.3 and 19.4 respectively. We conclude the chapter, ► Sect. 19.5, with an associative memory network.