Foreground process' fitness function
where: | l(h,a) | = |
number of games lost by each human opponent
h
versus each agent
ap(h,a)
| =
|
total number of games played,
a
versus
h | l(h)
| =
|
total number of games lost by
h | p(h)
| =
|
total number of games played by
h | |