高次ランダム神経回路網のパターン間の距離の変換特性

DOI Web Site Web Site 参考文献15件

書誌事項

タイトル別名
  • On Transformation of Distance between Patterns for Higher Order Random Neural Networks with Analog State
  • コウジ ランダム シンケイ カイロモウ ノ パターン カン ノ キョリ ノ ヘ

この論文をさがす

抄録

Many studies have been done with higher order neural networks. They include many studies from theory to application. With them, studies which have shown the superiority of higher order neural networks in combinatorial optimization problems, pattern classification, associative memory, and have given construction methods of effective networks in order to control the growth of the number of weights in using multi-layered higher order neural networks have been made. Although there are many applied studies as these, little is known about theoretical studies like macroscopic properties.<br>In the previous paper, we have shown transformation of distance between two patterns for higher order random neural networks (HORNNs) with the digital state {0, 1} models using statistical method. This paper describes transformation of distance between two patterns for HORNNs with the analog [0, 1] model using the same method and comparison between analog and digital higher order neural networks is made. Further, one example is given in order to show the superiority over the ability of them.

収録刊行物

参考文献 (15)*注記

もっと見る

キーワード

詳細情報 詳細情報について

問題の指摘

ページトップへ