摘要
在基于解释的机器学习问题上 ,近期提出的模糊模型FEBM(FuzzyExplanation -BasedModel)为模糊概念的识别和分类提供了一种很好的解决手段。在对该模型当对象的解释谓词在 [0 ,1]上取确值的情况时 ,计算“对象属于概念C的真值”的公式进行适当调整的基础上 ,结合神经网络可以用于模式识别和分类的特点 ,提出了一种基于模糊神经网络和FEBM的模糊概念识别方法。实验表明 ,该方法是有效的和可行的 ,是关于该模型应用的一个极为有意义的尝试。
In explanation-based learning area, FEBM (fuzzy explanation-based model), which was proposed by some scholars recently, supplies an important method to the identification and classification of fuzzy concepts. As to the formula calculating 'the true value of objects belonging to a concept C'in the model, some adjustments are given in this paper. Basing upon FEBM and neural network, which has been widely applied in pattern recognition and classification these years, a kind of fuzzy neural network used to identify fuzzy concepts is proposed here. Experiments show that it is efficient and applicable, and is also an important try about the application of FEBM.
出处
《微机发展》
2002年第4期25-27,共3页
Microcomputer Development
基金
安徽省教育厅科研经费资助项目 (2 0 0 0J1192 )
关键词
FEBM
机器学习
模糊神经网络
模糊概念识别
fuzzy neural network
fuzzy explanation-based model
fuzzy explanation set
explanation-based learning