PowerEvaluationBALD: Efficient Evaluation-Oriented Deep (Bayesian) Active Learning with Stochastic Acquisition Functions
This work addresses the computational bottleneck in deep Bayesian active learning for researchers and practitioners by enabling larger acquisition batch sizes with reduced computational demands.
This paper introduces PowerEvaluationBALD, a new acquisition function for deep Bayesian active learning that incorporates an evaluation set of unlabeled data and uses stochastic acquisition functions with importance sampling to reduce computational requirements. The method achieves performance on par with BatchEvaluationBALD, outperforming BatchBALD on Repeated MNIST, while significantly reducing computational costs.
We develop BatchEvaluationBALD, a new acquisition function for deep Bayesian active learning, as an expansion of BatchBALD that takes into account an evaluation set of unlabeled data, for example, the pool set. We also develop a variant for the non-Bayesian setting, which we call Evaluation Information Gain. To reduce computational requirements and allow these methods to scale to larger acquisition batch sizes, we introduce stochastic acquisition functions that use importance sampling of tempered acquisition scores. We call this method PowerEvaluationBALD. We show in a few initial experiments that PowerEvaluationBALD works on par with BatchEvaluationBALD, which outperforms BatchBALD on Repeated MNIST (MNISTx2), while massively reducing the computational requirements compared to BatchBALD or BatchEvaluationBALD.