Universum Learning for SVM Regression
This work addresses regression tasks by incorporating a priori knowledge through Universum data, but it appears incremental as it extends an existing concept to a new problem type.
The paper tackled the problem of extending Universum learning to regression by proposing a Universum-SVM formulation that incorporates additional data samples from the same domain but different distribution, with empirical comparisons showing its utility.
This paper extends the idea of Universum learning [18, 19] to regression problems. We propose new Universum-SVM formulation for regression problems that incorporates a priori knowledge in the form of additional data samples. These additional data samples or Universum belong to the same application domain as the training samples, but they follow a different distribution. Several empirical comparisons are presented to illustrate the utility of the proposed approach.