|
|
|
|
|
|
by Dn_Ab
4209 days ago
|
|
|
In other settings, a simple approximation of what Evolution does will balance your portfolio and find Nash Equilibriums for zero sum games much more quickly than linear programming (query: multiplicative update evolution). Having to wait for things to die can be restrictive (but genomes are declarative, what does that buy?). Roughly, with what amounts to the assumption that Lamarckian evolution is more or less not true, you can characterize evolution's learning power as almost as good as PAC (query: statistical query learning evolution). See alsos: http://vserver1.cscs.lsa.umich.edu/~crshalizi/weblog/601.htm... http://arxiv.org/abs/1211.4037: Natural selection. V. How to read the fundamental equations of evolutionary change in terms of information theory |
|
|