一种基于代价敏感学习的范例推理方法及其应用研究

Research on a case-based reasoning method using cost-sensitive learner and its applications

提出一种基于代价敏感学习的范例推理方法,可以对大规模、高维数据进行分类和预测。该算法在分类的同时,不断调整数据属性项权重,以减少由分类引起的误分代价。在某入侵检测数据分析中取得了较好的结果。

This paper proposed a case-based reasoning method using cost-sensitive learning that can classify and forecast large-scale and high dimension data. While classifying, this method adjust attribute weight constantly, in order to reduce the misclassification cost. The method has made a better result on some intrusion detection research.