中文摘要：

针对一类分类马氏椭球学习机（Mahalanobis ellipsoidal learning machine for one class classification，MELM）方法中选取参数C比较困难的问题，提出一种改进的方法v—MELM。这种方法通过引入一个具有明确物理意义的参数v，即v是超椭球外部的样本点数（野点数）占总样本点数的份额的上界，是支持向量的个数所占总样本点数的份额的下界，使参数可以灵活地根据实际问题的精度要求来选取，从而可以快速选取最有效的参数。

英文摘要：

If using Mahalanobis ellipsoidal learning machine （MELM） for one class classification, the difficulty in selecting the most effective error penalty C isn＇t resolved. A v-MELM classification method is proposed by improving upon MELM. This method provides the facility to counter these effects by introducing a parameter v which has specific meanings that represent the upper bound for the fraction of data vectors that lie outside the hyper-ellipsoid, i.e., the fraction of data vectors that can be outliers or anomalies, and the lower bound for the fraction of support vectors of whole data sets. As such this bound enable the training of machines with specific requirements and the most effective parameter is selected.