Can Active Learning Experience Be Transferred?
This addresses the challenge of reducing labeling effort in machine learning by enabling experience transfer, though it is incremental as it builds on existing active learning strategies.
The paper tackles the problem of transferring active learning experience across datasets, proposing a model that aggregates strategies with linear weights updated via contextual bandit algorithms. Empirical results show the learned experience is competitive on single datasets and improves performance when transferred to new tasks.
Active learning is an important machine learning problem in reducing the human labeling effort. Current active learning strategies are designed from human knowledge, and are applied on each dataset in an immutable manner. In other words, experience about the usefulness of strategies cannot be updated and transferred to improve active learning on other datasets. This paper initiates a pioneering study on whether active learning experience can be transferred. We first propose a novel active learning model that linearly aggregates existing strategies. The linear weights can then be used to represent the active learning experience. We equip the model with the popular linear upper- confidence-bound (LinUCB) algorithm for contextual bandit to update the weights. Finally, we extend our model to transfer the experience across datasets with the technique of biased regularization. Empirical studies demonstrate that the learned experience not only is competitive with existing strategies on most single datasets, but also can be transferred across datasets to improve the performance on future learning tasks.