Merge evolved artificial neural network strategy #773
Conversation
| return D | ||
|
|
||
|
|
||
| class EvolvedANN(ANN): |
There was a problem hiding this comment.
Can we have more of a docstring here please.
-
Saying that it's based on a neural net etc.
-
Include:
Names: - Evolved ANN: : Original name by Marc Harper/Marting Jones
There was a problem hiding this comment.
Perhaps also include a reference (in bibliography.rst) to a good base text on neural nets? (Perhaps.)
There was a problem hiding this comment.
Yes to docstring, I think people should just google about neural networks.
| input_values = 17 | ||
| hidden_layer_size = 10 | ||
|
|
||
| test_weights = [0.19789658035994948, -5575.476236516673, 0.1028948855131803, 0.7421752484224489, |
There was a problem hiding this comment.
Is test_weights the right name for this?
| class TestEvolvedANNvsCooperator(TestHeadsUp): | ||
| def test_rounds(self): | ||
| self.versus_test(axelrod.EvolvedANN(), axelrod.Cooperator(), | ||
| [C, D, D, C, D], [C] * 5) |
There was a problem hiding this comment.
Could we throw in a couple more? (Just to ensure we have some more variety).
| opponent_previous2_d = 1 if opponent.history[-2] == D else 0 | ||
|
|
||
| # Remaining Features | ||
| turns_remaining = self.match_attributes['length'] - len(self.history) |
There was a problem hiding this comment.
Can we add a test for the case of a match when the match length is unknown? (Just to make sure this doesn't fall over).
There was a problem hiding this comment.
The heads up tests don't supply the match length.
There was a problem hiding this comment.
We should train a separate strategy that only uses the round number though.
No description provided.