2024-03-29T21:13:22Zhttps://ipsj.ixsq.nii.ac.jp/ej/?action=repository_oaipmhoai:ipsj.ixsq.nii.ac.jp:000389872024-03-29T05:26:34Z01164:03206:03317:03318
パターン認識の数学的理論 第XV部 パターンの構造的類似性をもたらす千種類の収縮写像A MATHEMATICAL THEORY OF RECOGNIZING PATTERNS PART XV FOUR KINDS OF CONTRACTION - MAPPINGS WHICH CAN REPRODUCE A STRUCTURAL SIMILARITY OF PATTERNSjpnhttp://id.nii.ac.jp/1001/00038987/Technical Reporthttps://ipsj.ixsq.nii.ac.jp/ej/?action=repository_action_common_download&item_id=38987&item_no=1&attribute_id=1&file_no=1Copyright (c) 1989 by the Information Processing Society of Japan文教大学湘南キャンパス情報学部情報システム学科鈴木, 昇一We make four kinds of contraction-mapping which provide one another an approximation of a pattern. This four mappings make a respective approximation with a linear combination of primitive patterns and differ in methods of determining coefficients (adaptive or non-adaptive and binary or continuous coefficients) of the linear combination. The four mappings are provided in order to realize the functions of neural networks i.e. the Hopfield associative memories and the multilayer Perceptrons in the framework of the mathematical theory of recognizing patterns.We make four kinds of contraction-mapping which provide one another an approximation of a pattern. This four mappings make a respective approximation with a linear combination of primitive patterns and differ in methods of determining coefficients (adaptive or non-adaptive and binary or continuous coefficients) of the linear combination. The four mappings are provided in order to realize the functions of neural networks, i.e. the Hopfield associative memories and the multilayer Perceptrons in the framework of the mathematical theory of recognizing patterns.AN10100541情報処理学会研究報告グラフィクスとCAD(CG)1989109(1989-CG-042)181989-12-142009-06-30