模糊系统与数学
模糊繫統與數學
모호계통여수학
FUZZY SYSTEMS AND MATHEMATICS
2004年
2期
62-67
,共6页
归结%神经网络%子句集%完备性
歸結%神經網絡%子句集%完備性
귀결%신경망락%자구집%완비성
Resolution%Neural Network%Clause Set%Completeness Theorem
给出基于神经网络的归结方法.首先将子句集S表示为δ形式,并且用算子对(·,+))引入两种类型的神经元; 然后用这两种神经元构造子句集S的神经网络结构; 而后给出基于子句集的神经网络的归结算法; 最后证明了该算法的完备性,并用实例进行了验证.
給齣基于神經網絡的歸結方法.首先將子句集S錶示為δ形式,併且用算子對(·,+))引入兩種類型的神經元; 然後用這兩種神經元構造子句集S的神經網絡結構; 而後給齣基于子句集的神經網絡的歸結算法; 最後證明瞭該算法的完備性,併用實例進行瞭驗證.
급출기우신경망락적귀결방법.수선장자구집S표시위δ형식,병차용산자대(·,+))인입량충류형적신경원; 연후용저량충신경원구조자구집S적신경망락결구; 이후급출기우자구집적신경망락적귀결산법; 최후증명료해산법적완비성,병용실례진행료험증.
In this paper, we give a method of resolution by neural network. First, we express the clause set S in σ-form. Then, introduce two kinds of neurons by a couple of operators (·,+) and get NN structure of σ-form clause set S. After that resolution algorithm based on the NN is obtained. In the end, prove the complete theorem of the resolution algorithm and test it by example.