DOC3-Deep One Class Classification using Contradictions
This work addresses one-class classification, a domain-specific problem, with an incremental improvement by incorporating contradictions into existing methods.
The paper tackles the problem of one-class classification by introducing learning from contradictions (Universum learning) into deep models, showing that the proposed DOC3 algorithm achieves lower generalization error and outperforms baseline methods on real-world datasets.
This paper introduces the notion of learning from contradictions (a.k.a Universum learning) for deep one class classification problems. We formalize this notion for the widely adopted one class large-margin loss, and propose the Deep One Class Classification using Contradictions (DOC3) algorithm. We show that learning from contradictions incurs lower generalization error by comparing the Empirical Rademacher Complexity (ERC) of DOC3 against its traditional inductive learning counterpart. Our empirical results demonstrate the efficacy of DOC3 compared to popular baseline algorithms on several real-life data sets.