层次分析法和灰色聚类分析法在绩效评估中的应用

《层次分析法和灰色聚类分析法在绩效评估中的应用》由会员上传分享，免费在线阅读，更多相关内容在应用文档-天天文库。

1、层次分析法和灰色聚类分析法在绩效评估中的应用施狄峰摘要　　绩效考核的评估是帮助企业维持和提高生产力、实现企业经营目标的手段之一，它一个复杂的大系统，一般企业的绩效评估是建立在关键考核指标得分乘以权系数的线性关系的基础上，但如果有两个下属分公司考核得分分别是97分和94分，究竟它们都属于优，还是一个是优、一个是良，原先的方法显然无法判断。笔者运用运筹学决策分析法的层次分析法和灰色系统理论的灰聚类法两种方法对绩效加以评估，能将被考核企业的经营情况很清楚地区分开来，分类排序出来。关键词　绩效评估　层次分

2、析法灰色聚类分析法　设以某公司下属11个分公司绩效考核情况数据为例，记为;并选取经营效绩考核中三个指标记为。一、用层次分析法：1、权重设置：根据三个指标的权重系数A1，A2，A3两两比较，得到数值aij,其定义和解释见表1，得出表2所示系数。相对重要程度定义解释1同等重要目标i和j同样重要3略微重要目标i比j略微重要5相当重要目标i比j重要7明显重要目标i比j明显重要9绝对重要目标i和j绝对重要2、4、6、8介于两相邻重要程度间表1考核指标比值考核指标11.1330.8912.670.330.38

3、1表2得到矩阵A=(aij)3×3矩阵A为经营效绩的判断矩阵。A=相应的特征向量为：B3=(0.450.400.15)T得出3个考核指标权重分别为0.45、0.40、0.1582、类似地根据表3可用特征向量法求下属11个分公司相对于上述3个指标中每一个的权系数。成对比较的结果如表4、表5、表6所示：下属分公司代号A1.020.971.05B1.161.101.07C0.900.791.01D1.000.981.13E1.051.061.26F0.790.641.10G0.961.001.06H0.

4、951.111.07I1.001.140.92J0.820.710.98K0.891.001.11表3指标：ABCDEFGHIJKA1.000.881.131.020.971.291.061.071.021.241.15B1.141.001.291.161.101.471.211.221.161.411.30C0.880.781.000.900.861.140.940.950.901.101.01D0.980.861.111.000.951.271.041.051.001.221.12E1.030

5、.911.171.051.001.331.091.111.051.281.18F0.770.680.880.790.751.000.820.830.790.960.89G0.940.831.070.960.911.221.001.010.961.171.08H0.930.821.060.950.901.200.991.000.951.161.07I0.980.861.111.000.951.271.041.051.001.221.12J0.800.710.910.820.781.040.850.8

6、60.821.000.92K0.870.770.990.890.851.130.930.940.891.091.00表4指标：ABCDEFGHIJKA1.000.881.230.990.921.520.970.870.851.370.97B1.131.001.391.121.041.721.100.990.961.551.10C0.810.721.000.810.751.230.790.710.691.110.79D1.010.891.241.000.921.530.980.880.861.380

7、.98E1.090.961.341.081.001.661.060.950.931.491.06F0.660.580.810.650.601.000.640.580.560.900.64G1.030.911.271.020.941.561.000.900.881.411.00H1.141.011.411.131.051.731.111.000.971.561.11I1.181.041.441.161.081.781.141.031.001.611.148J0.730.650.900.720.671

8、.110.710.640.621.000.71K1.030.911.271.020.941.561.000.900.881.411.00表5指标：ABCDEFGHIJKA1.000.981.040.930.830.950.990.981.141.070.95B1.021.001.060.950.850.971.011.001.161.090.96C0.960.941.000.890.800.920.950.941.101.030.91D1.081.061.121.000.901.0