Multi-task and Lifelong Learning of Kernels
This addresses the challenge of efficient kernel selection for SVM classification in multi-task settings, though it appears incremental as it builds on existing generalization bounds and kernel learning frameworks.
The paper tackles the problem of learning kernels for SVM classification in multi-task and lifelong learning scenarios, showing that solving multiple related tasks simultaneously improves over single-task learning. The result demonstrates that as the number of tasks grows, the complexity of learning converges to that of having an optimal kernel provided, with the overhead vanishing under mild conditions.
We consider a problem of learning kernels for use in SVM classification in the multi-task and lifelong scenarios and provide generalization bounds on the error of a large margin classifier. Our results show that, under mild conditions on the family of kernels used for learning, solving several related tasks simultaneously is beneficial over single task learning. In particular, as the number of observed tasks grows, assuming that in the considered family of kernels there exists one that yields low approximation error on all tasks, the overhead associated with learning such a kernel vanishes and the complexity converges to that of learning when this good kernel is given to the learner.