论文部分内容阅读
在运用RSR法解决实际问题时,第一步就要编秩计算RSR值。统计学家田凤调以前所述的编秩方法是:在m个指标中,首先要确定各指标是属于高优、偏高优、不分、偏低优、低优等中的哪一类,然后再分别编秩。若有n个样品,则不分指标按C_0=1+n/2编秩;高优指标按数值从小到大依次编1,2,…n;偏高优=1/2(高优+不分);稍高优=1/2(偏高优+不分);低优指标按数值从大到小依次编1,2,…n;偏低优=1/2(低优+不分);稍低优=1/2(偏低优+不分);…
When using RSR to solve practical problems, the first step is to rank and rank RSR values. Statistician Tian Feng tune ranking method as described above is: in the m indicators, we must first determine the indicators are high, high and low, regardless of low, low and excellent in what kind of, and then Then rank separately. If there are n samples, regardless of the index by C_0 = 1 + n / 2 rank; high-quality index according to the value of small to large in turn compiled 1,2, ... n; high high = 1/2 (high + Points); slightly higher = 1/2 (high high + no distinction); low high index according to the value of descending order 1,2, ... n; low low = 1/2 ); Slightly lower excellent = 1/2 (low excellent + indifferent); ...