An l1 Regularization Framework for Optimal Rule Combination
Published on Oct 20, 20092781 Views
In this paper l1 regularization is introduced into relational learning to produce sparse rule combination. In other words, as few as possible rules are contained in the final rule set. Furthermore, we