论文部分内容阅读
本文根据风险投资的特点,建立了一个多属性决策问题的模型。利用“奖优罚劣”的转换办法初始化原始决策矩阵,又根据fuzzy相似类、变精度粗糙集和近似分类质量等转化和处理得到属性约简的信息系统,从而求出重要属性的客观权重来评价各个投资方案的优劣。最后利用属性权重完全未知且属性值为实数的多属性风险投资决策案例来验证模型的有效性。
According to the characteristics of venture capital, this paper establishes a model of multi-attribute decision making problem. The original decision-making matrix is initialized by the conversion method of “rewards and penalties”, and then the information system of attribute reduction is obtained through the conversion and processing of fuzzy similar categories, variable precision rough sets and approximate classification quality, so as to obtain objective attributes of important attributes The weight to evaluate the merits of various investment programs. Finally, the validity of the model is verified by using multi-attribute VC decision cases whose attribute weights are completely unknown and whose attribute values are real numbers.