We transform small and middle to the big size indicators in following part.
-
For the big size indicators
In order to remove the effect caused by the dimension and magnitude, the following function is used to standardize.
Where,
-
denotes the raw value of the
country in
indictor.
-
denotes the normalized value of the
country in
indictor.
Taking "Population ages 0-14" for an example,we can get the normalized data.
Table 2 The indicator- population ages 0-14 is normalized
Country | America | Japan | Brazil | Philippines | Afghanistan |
Population ages 0-14 | 0.21 | 0 | 0.32 | 0.62 | 1 |
-
For the middle size indicators
The middle size indicator is transformed to the big size indictor by the formula.
Where
Taking "Ratio of male to female" as an example,the normalized data is gotten:
Table 3 The ratio of male to female
Country | America | Japan | Brazil | Philippines | Afghanistan |
The ratio of male to female | 0.94 | 0.95 | 0.87 | 0.92 | 0.67 |
-
For the small size indicators
Similarly, the small size indicator is transformed by the following formula.
Taking "Population ages 65 and above" as an example, we can get the normalized data.
Table 4 Population ages 65 and above
Country | America | Japan | Brazil | Philippines | Afghanistan |
Population ages 65 and above | 0.48 | 0 | 0.74 | 0.91 | 1 |
将区间型化为极大型
对某个区间型指标x,则
其中为 x 的最佳稳定区间,
,M 和 m 分别为 x 可能取值的最大值和最小值。
clc,clear;
%----------
% 对象 对象 对象
% 指标
% 指标
% -----------%
% x=[1 2 3 5 6;1 20 3 5 6];
x=load('data.txt');
%输出矩阵y
m=size(x,1);
n=size(x,2);
y=ones(m,n);
xy=minmax(x);
for i=1:m
for j=1:n
y(i,j)=(x(i,j)-xy(i,1))/(xy(i,2)-xy(i,1));
end
end
% function y=mscor(x)
clc,clear;
%----------
% 对象 对象 对象
% 指标
% 指标
% -----------%
x=[1 2 3 5 6;1 20 3 5 6];
% x=load('data.txt');
%输出矩阵y
m=size(x,1);
n=size(x,2);
y=ones(m,n);
xy=minmax(x);
for i=1:m
mid=(xy(i,2)+xy(i,1))/2;
for j=1:n
if x(i,j)>=xy(i,1)&&x(i,j)<=mid
y(i,j)=2*(x(i,j)-xy(i,1))/(xy(i,2)-xy(i,1));
else if x(i,j)<=xy(i,2)&&x(i,j)>=mid
y(i,j)=2*(xy(i,2)-x(i,j))/(xy(i,2)-xy(i,1));
end
end
end
end