信用评价论文：基于糙集与支持向量机的电力行业信用评价.docVIP

信用评价论文：基于粗糙集与支持向量机的电力行业信用评价 【中文摘要】随着电力体制改革逐步深入,我国电力供需情况发生很大变化,电力行业信用是电力企业面临的重要问题,不仅影响到企业的经营状况和企业形象,同时也关系到企业的生存和发展,因此有必要对电力行业的信用进行综合评价,从而规避电力企业因诚信问题所带来的风险。对电力客户的需求来讲,较为准确地确定电力企业的信用水平是十分重要的。本文根据目前电力行业信用评估的现状,确立了电力行业信用评价指标体系,针对指标数量较多的特点,提出了一种基于粗糙集和支持向量机的信用评价模型。借助粗糙集的属性约简算法实现评价指标约简;利用支持向量机的回归算法,构造出具有较好的泛化能力及抑制噪声干扰的回归模型,对实际数据进行指标筛选、权重计算及综合评估,确定电力企业信用评价的综合值,得到各电力企业优化的信用评价次序排名。并与单一支持向量机的模型相比较,验证了所选方法的有效性,研究结果表明该模型切实可行,效果良好。最后,针对目前我国电力企业信用评价的现状和存在的问题,提出了一些针对性的建议。 【英文摘要】With the gradual deepening of power system, power supply and demand situation in China has greatly changed, the credit of electricity power industry enterprises is a important issues which they are facing, not only affect the business operations and corporate image, but also related to survival and development of enterprises, so it is necessary to evaluate the credit of the power industry, thereby circumventing the problem of power companies for credit risks. In terms of customer demand for electricity, it is very important to accurately determine the level of power companies抍redit.This article based on the current status of the power industry抯 credit assessment, established credit evaluation index system of power industry. For the features of a large number of indicators, proposed based on rough sets and support vector machine credit evaluation model. Using rough sets of the attribute reduction algorithm to realize evaluation index reduction, Using support vector machine (SVM) is constructed, the regression algorithm has better generalization ability and the suppressing noise regression model, with real data index selection, weight calculation and comprehensive assessment, determine the electric power enterprise credit evaluation of comprehensive value, obtains each electric power enterprises optimized credit evaluation order rankings. And with the single support vector machine (SVM) model compared to verify the selected the effectiveness of the method, the results of the study show that the model is fea