An Unbiased Risk Estimator for Learning with Augmented Classes
This addresses the challenge of handling unseen classes in machine learning, particularly for applications with dynamic data, though it is incremental as it builds on prior geometric methods by adding theoretical guarantees.
The paper tackles the problem of learning with augmented classes (LAC), where new classes emerge during testing, by developing an unbiased risk estimator using unlabeled data to approximate the distribution of these classes, enabling a theoretically sound approach that adapts to changing environments.
This paper studies the problem of learning with augmented classes (LAC), where augmented classes unobserved in the training data might emerge in the testing phase. Previous studies generally attempt to discover augmented classes by exploiting geometric properties, achieving inspiring empirical performance yet lacking theoretical understandings particularly on the generalization ability. In this paper we show that, by using unlabeled training data to approximate the potential distribution of augmented classes, an unbiased risk estimator of the testing distribution can be established for the LAC problem under mild assumptions, which paves a way to develop a sound approach with theoretical guarantees. Moreover, the proposed approach can adapt to complex changing environments where augmented classes may appear and the prior of known classes may change simultaneously. Extensive experiments confirm the effectiveness of our proposed approach.