|
|
|
|
|
by hexomancer
3232 days ago
|
|
Thanks for the answer, I have edited my post and I have basically the same solution. (Only instead of removing the edges I increase their weight to infinity because I think the TSP problem is defined only on complete graphs) But as I have mentioned in my comment I feel like this solution is flawed. |
|
Hence, you can use bisection to compute the actual optimum, not involving epsilon at all.