Multi-Task Classification Hypothesis Space with Improved Generalization Bounds
This work addresses theoretical generalization guarantees for multi-task learning practitioners, though it appears incremental as it extends existing hypothesis spaces and bounds.
The paper tackles the problem of multi-task classification by proposing a new RKHS hypothesis space for vector-valued functions, deriving improved generalization bounds based on Empirical Rademacher Complexity, and validating these bounds with SVM-based experiments.
This paper presents a RKHS, in general, of vector-valued functions intended to be used as hypothesis space for multi-task classification. It extends similar hypothesis spaces that have previously considered in the literature. Assuming this space, an improved Empirical Rademacher Complexity-based generalization bound is derived. The analysis is itself extended to an MKL setting. The connection between the proposed hypothesis space and a Group-Lasso type regularizer is discussed. Finally, experimental results, with some SVM-based Multi-Task Learning problems, underline the quality of the derived bounds and validate the paper's analysis.