排序问题中的范数学习算法

Norm Learning Methods for Ranking

为了更直接地定义点对间的序关系,引入了排序问题的范数学习模型。该模型利用样本空间中的范数类描述样本间的不同序关系。借鉴范数学习的思想,得到了有效处理此类排序学习问题的贪婪学习算法。实验结果表明,此类算法排序效果良好。

Ranking is a valuable problem in the field of machine learning. To directly measure the ranking relation between both elements, a novel norm learning ranking model is reported. The model uses different ranking relations of the norm functions defined on the sample space. By using the norm learning methods, an optimal learning algorithm is obtained. Experimental results show that the reported ranking algorithm is effective.