6234RE: Starting with Neat, the XOR problem, some questions
- Jan 6 7:00 AM
Ok, re-reading a few more time and trying to implement the logic I have found a flaw in my understanding or in the logic of the process.
For each member of the population you find the distance vs every other member of the population. If the distance is < x then then the populations belong in the same species.
So lets take the following 3 members of the population A, B and C and calculate the distances.
A <=> B, distance is OK
A <=> C, distance is NOT OK
B <=> C, distance is OK
From the document "A given genome g in the current generation is placed in the first species in which g is compatible with the representative genome of that species."
By that logic I start with A. A is the first member so I start a new Species with A in it. A is compatible with B so A and B are in the same species. So, after looking at A and B I have just 1 species.
When I get to C, I look for the first species which contains a member that is compatible with C. B is compatible with C and the first species that contains B is the first species, so I should put C into the first species. But this wrong, because the first species contains A and A is not compatible with C!
I think I will stop for today and wait for some clarification.
- << Previous post in topic Next post in topic >>