Open wazaahhh opened 8 years ago
I think there is a problem with arbitrarily setting np.array([1]*8)/8.0, because as we found, it makes the Bayesian learner start with a edge, compared to humans. It would be better to start with np.random.rand(8), i.e., 8 random values.
I think there is a problem with arbitrarily setting np.array([1]*8)/8.0, because as we found, it makes the Bayesian learner start with a edge, compared to humans. It would be better to start with np.random.rand(8), i.e., 8 random values.