Niko/ActiveBOToytask
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Compare commits

base: Niko:432978a1220496575f81c313f7e7f1292bc1d650
Branches Tags
Niko:master
..
compare: Niko:3f9b12e7f2f547ed1b61fa7d9584b92d0d37c443
Branches Tags
Niko:master

No commits in common. "432978a1220496575f81c313f7e7f1292bc1d650" and "3f9b12e7f2f547ed1b61fa7d9584b92d0d37c443" have entirely different histories.
432978a122 ... 3f9b12e7f2

1 changed files with 10 additions and 11 deletions
Show Stats Expand all files Collapse all files

21
AcquistionFunctions/PreferenceExpectedImprovement.py
View File

@ -28,7 +28,7 @@ class PreferenceExpectedImprovement:
while samples.shape[0] < self.nr_samples: while samples.shape[0] < self.nr_samples:
# sample from the multi variate gaussian distribution # sample from the multi variate gaussian distribution
sample = self.rng.multivariate_normal( sample = self.rng.multivariate_normal(
self.proposal_model_mean, self.proposal_model_mean,
self.proposal_model_covariance self.proposal_model_covariance
) )
@ -43,21 +43,20 @@ class PreferenceExpectedImprovement:
pass pass
def update_user_preference_model(self, preferred_input, preference_array): def update_user_preference_model(self, preferred_input, preference_array):
# Update mean to reflect preferred input # Update mean to reflect preferred input
self.user_model_mean = preferred_input self.user_model_mean = preferred_input
initial_variance = np.ones((self.nr_dims, )) * self.initial_variance initial_variance = np.ones((self.nr_dims, )) * self.initial_variance
reduced_variance = initial_variance / 10.0 reduced_variance = initial_variance / 10.0
variances = np.where(preference_array, reduced_variance, initial_variance) variances = np.where(preference_array, reduced_variance, initial_variance)
self.user_model_covariance = np.diag(variances) self.user_model_covariance = np.diag(variances)
def update_proposal_model(self, alpha=0.5): def update_proposal_model(self, alpha=0.5):
# Update proposal model to be a weighted average of the current proposal model and the user model # Update proposal model to be a weighted average of the current proposal model and the user model
self.proposal_model_mean = alpha * self.proposal_model_mean + (1 - alpha) * self.user_model_mean self.proposal_model_mean = alpha * self.proposal_model_mean + (1 - alpha) * self.user_model_mean
self.proposal_model_covariance = alpha * self.proposal_model_covariance + (1 - alpha) * self.user_model_covariance self.proposal_model_covariance = alpha * self.proposal_model_covariance + (1 - alpha) * self.user_model_covariance
if __name__ == '__main__': if __name__ == '__main__':
acquisition = PreferenceExpectedImprovement(10, 2, -1.0, 1.0) acquisition = PreferenceExpectedImprovement(10, 2, -1.0, 1.0)