Improving Active Learning with a Bayesian Representation of Epistemic Uncertainty
This work addresses the problem of improving active learning efficiency for researchers and practitioners by offering a novel uncertainty representation, though it appears incremental as it builds on existing theories.
The paper tackled the challenge of distinguishing epistemic and aleatoric uncertainty in active learning by proposing a combination of probability and possibility theories, leading to new strategies that showed strong performance in GP-based classification tasks on simulated and real datasets.
A popular strategy for active learning is to specifically target a reduction in epistemic uncertainty, since aleatoric uncertainty is often considered as being intrinsic to the system of interest and therefore not reducible. Yet, distinguishing these two types of uncertainty remains challenging and there is no single strategy that consistently outperforms the others. We propose to use a particular combination of probability and possibility theories, with the aim of using the latter to specifically represent epistemic uncertainty, and we show how this combination leads to new active learning strategies that have desirable properties. In order to demonstrate the efficiency of these strategies in non-trivial settings, we introduce the notion of a possibilistic Gaussian process (GP) and consider GP-based multiclass and binary classification problems, for which the proposed methods display a strong performance for both simulated and real datasets.