CAJ | 학술논문

在基于粗糙集理论或扩展粗糙集理论的不确定性决策中,仅由数据集无法获得决策表,同时由于决策属性呈现模糊性,以致于难以获取概率决策规则。针对这一类决策问题,基于灰色定权聚类与优势粗糙集变精度方法,本文构建了一种的杂合决策方法。该方法首先利用中心点三角白化权函数的灰色定权聚类方法将包含不同量纲指标的知识表示系统生成多属性决策表,而后运用变精度粗糙模糊集进行决策分析,生成概率决策规则,最后以区域关键技术的选择为例,表明该模型的可行性与实用性。
재기우조조집이론혹확전조조집이론적불학정성결책중,부유수거집무법획득결책표,동시유우결책속성정현모호성,이치우난이획취개솔결책규칙。침대저일류결책문제,기우회색정권취류여우세조조집변정도방법,본문구건료일충적잡합결책방법。해방법수선이용중심점삼각백화권함수적회색정권취류방법장포함불동량강지표적지식표시계통생성다속성결책표,이후운용변정도조조모호집진행결책분석,생성개솔결책규칙,최후이구역관건기술적선택위례,표명해모형적가행성여실용성。
Facing today＇ s fiercely competitive market, businesses or individuals often encounter complex decision making concerning uncertainty, and need to analyze and deal with various uncertainties such as randomness, fuzziness, preferences, roughness and greyness. However, only a theory or method based on soft computing technology can handle many uncertain decision-making problems. Some soft computing technologies, such as rough set theory, the grey system theory, fuzzy set theory and their complementary advantages, can be used to flexibly deal with uncertainties in real life. A new mathematic tool is needed to deal with inaccurate, inconsistent and incomplete information. The rough set theory possesses its advantages, disadvantages and the scope of application, and has difficulties in dealing with some uncertain decision-making problems. Therefore, according to the variety of uncertainties in the practical application of knowledge representation systems the advantages of the rough set theory with those of other soft technology theories can be integrated to build a more powerful hybrid method of soft decision making, broaden the scope of the rough set theory application, and provide a wide range of scientific and standard methods for making decisions about uncertainties. This paper proposes a novel hybrid probabilistic decision approach based on the center triangle whitenization weight function with fixed weight and variable precision rough fuzzy set. This approach can help solve decision-making problems when decision tables cannot be generated from the data, decision attributes possess fuzziness, and probabilistic decision rules can＇ t be obtained in many uncertain practical decisions based on rough set models or extensions of rough set model. The proposed method is applied to the selection of regional key technology, and the results show that the method can be feasible and practical. In the first part, some problems, such as the grey comprehensive clustering based on the center triangle whitenization weight function, build multiple attribute decision table, variable precision rough fuzzy set, and probabilistie decision rules. The grey comprehensive clustering is based on the center triangle whitenization weight function with fixed weight and equal weight. The grey comprehensive clustering based on the center triangle whitenization weight function with fixed weight is taken as the pre-processing tool for variable precision rough fuzzy set to convert multiple attribute information table into multiple attribute decision table. In view of the fact that the approximation space is clear and decision attribute value is fuzzy, variable precision rough fuzzy set is proposed. The clustering coefficients are used as the fuzzy membership and the probabilistic decision analysis is made to deal with the information system containing fuzzy data. The probabilistic decision rules are derived. In the second part, the indirect model validates the proposed method＇ s feasibility and practicality. The hybrid probabilistic decision approach of grey clustering based on the center triangle whitenization weight function with fixed weight and variable precision rough fuzzy set is applied to the selection of regional key technology in Jiangsu. The results show that the method can realize the rapidreduction and possess those abilities of improving the classification accuracy and the fault tolerance by adjusting the parameter/3. In summary, the hybrid probabilistic decision approach of grey clustering based on the center triangle whitenization weight function with fixed weight and variable precision rough fuzzy set can be used to deal with decision problems that are impossible to generate decision tables from data set. The approximation space is clear and decision attribute value is fuzzy. Thus, it can help improve the quality and efficiency of decision-making and achieve decision-making goals.