Bridging Ordinary-Label Learning and Complementary-Label Learning
This work addresses a theoretical gap in machine learning by integrating complementary-label learning with standard supervised learning, which is incremental in nature.
The paper tackles the problem of unifying ordinary-label and complementary-label learning by proposing a framework that bridges these two supervised learning approaches, deriving classification risk and error bounds for loss functions with specific properties.
A supervised learning framework has been proposed for the situation where each training data is provided with a complementary label that represents a class to which the pattern does not belong. In the existing literature, complementary-label learning has been studied independently from ordinary-label learning, which assumes that each training data is provided with a label representing the class to which the pattern belongs. However, providing a complementary label should be treated as equivalent to providing the rest of all the labels as the candidates of the one true class. In this paper, we focus on the fact that the loss functions for one-versus-all and pairwise classification corresponding to ordinary-label learning and complementary-label learning satisfy certain additivity and duality, and provide a framework which directly bridge those existing supervised learning frameworks. Further, we derive classification risk and error bound for any loss functions which satisfy additivity and duality.