without taking into account the model’s uncertainty can be very misleading, particularly when the estimated model is far from the true model. A better strategy could
be to consider model uncertainty by utilizing the model distribution for selecting
training input points [25]. This would allow for adjusting the decision boundary
more effectively since decision boundaries other than the estimated one (i.e. horizontal line) would be considered for selecting the training input points. This idea is
applied to probabilistic models in [25] as follows. The usefulness of the candidate
training input point is measured based on how much it allows adjusting the model’s
parameters θ u towards the optimal model parameters θ u∗: