Searched refs:selected_policy (Results 1 – 1 of 1) sorted by relevance
419 def update_metrics(self, access_time, selected_policy): argument426 policy_name = self.policy_names[selected_policy]1034 reward = self.policies[selected_policy].generate_reward(key)1036 self._as[selected_policy] += reward1037 self._bs[selected_policy] += 1 - reward1038 return selected_policy1090 selected_policy = p.argmax()1093 self.A[selected_policy] += np.outer(x_i, x_i)1094 self.b[selected_policy] += reward * x_i1095 self.A_inv[selected_policy] = np.linalg.inv(self.A[selected_policy])[all …]