how can i formulate this fitness function ?

3 次查看（过去 30 天）

Ameer Kadhim 2018-3-10

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/387492-how-can-i-formulate-this-fitness-function

评论： Star Strider 2018-3-10

hello everybody...

my problem is : i have matrix consist of 4 columns and 200 rows , the fist column represent the index , the second column contain the values of weights ( in kg) , third column contain the values of length(in meter) , the fourth column contains the values of age(in days) . how i can formulate this problem in genetic algorithm (e.g NSGA-II)in matlab , such that finally i want the number vector or pareto-front that contain (low value of weight, high value of length, less value of age ) simultaneously. for example:

1 : 20 70 45

2: 49 10 67

3: 100 15 46

4: .... and so on .

And what is form of the fitness function? Please help me...

With sincere thanks and appreciation.....

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

采纳的回答

Star Strider 2018-3-10

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/387492-how-can-i-formulate-this-fitness-function#answer_309413

在 MATLAB Online 中打开

The function you want is gamultiobj (link).

I am not certain what you want to do, so I leave the design of the fitness function to you.

Something like this could work:

M = [(1:25)' randi(100, 25, 3)];                                            % Create Data Matrix
fitfcn = @(x) [x(1).*min(M(2)), x(2).*max(M(3)), x(3).*min(M(4))];
X = gamultiobj(fitfcn, 3)
Mval = [X(:,1).*min(M(:,2)), X(:,2).*max(M(:,3)), X(:,3).*min(M(:,4))];
figure(1)
scatter3(Mval(:,1), Mval(:,2), Mval(:,3), 'p')
grid on
figure(2)
stem3(Mval(:,1), Mval(:,2), Mval(:,3), 'p', 'Filled')
grid on

6 个评论
显示 4更早的评论隐藏 4更早的评论

Ameer Kadhim 2018-3-10

Do you know someone who can help me please. I'm tired. I did not find the solution ..

Star Strider 2018-3-10

There are probably people here on MATLAB Answers who have the expertise you need. That is the reason I suggested that you ask a new Question, this time describing the problem you actually want to solve.

请先登录，再进行评论。