Abstract: The data that Support Vector Machine （SVM） deals with are mostly precise values, but the SVM cannot be utilized when training samples involving in fuzzy information. Based on this, in response to the classification problem that input data are fuzzy numbers, a novel Fuzzy SVM （FSVM*） with defuzzification function is proposed. This algorithm constructs defuzzification function by taking the distance between fuzzy numbers as the metric for defuzzification of fuzzy numbers to convert fuzzy numbers into precise values, and classifies the fuzzy data by using defuzzification function and fuzzy SVM in combination at the same time. The experimental results show that the  model in this paper is more effective compared to the FSVDD* proposed by Forghani.

Key words: fuzzy support vector machine, fuzzy number, defuzzification function, distance

摘要： 支持向量机所处理的数据绝大多数是精确值，但当训练样本中含有模糊信息时，支持向量机将无能为力。基于此，针对输入数据是模糊数的分类问题，提出一种带有去模糊函数的模糊支持向量机（FSVM*）。该算法采用模糊数间的距离作为模糊数去模糊的度量，从而构造去模糊函数将模糊值转化为精确值，同时将去模糊函数与模糊支持向量机相结合完成模糊数据的分类。数值结果表明：相比Forghani提出的FSVDD*算法，该算法更有效。

关键词: 模糊支持向量机, 模糊数, 去模糊函数, 距离