How to calculate the crowding distance of the objective values ?

let think I have a matrix A ,After non dominated sorting I ranked them.I have 15 variables column number (1-15) and two objectives column number(16-17) based on their objective values i ranked them from 1 to N. Now i want to get the crowding distance for both the objective values .As per the formula

CD=(next objective value - Previous objective value )/(Fmax-Fmin)

How to solve this ?I can't understand it .

Answer 1

[Ayush Modi]

Hi,

I understand that you want to know how to implement the logic of calculating crowding distance for each objective parameter (Column 16 and 17).

You can achieve this by using “distancecrowding” inbuilt function in MATLAB Optimization Toolbox.

Alternatively, I found a solution posted in the community. You can refer to the following link:

https://www.mathworks.com/matlabcentral/fileexchange/65809-on-the-calculation-of-crowding-distance

Go to the functions tab at the top of the page. You will find two files – “CheckCrowdingDistance.m” and “non_domination_sort_mod.m”. Refer “non_domination_sort_mod.m” file for the implementation.

I have tried at my end, and both the solution is working. Hope this helps!