Fig. 5From: CLeaR: An adaptive continual learning framework for regression tasksA probability distribution of errors for the data in the novelty buffer (or training dataset) and the familiarity buffer (or test dataset). MSEmin indicates the minimum MSE after learningBack to article page