The Method of Sample Selection Based on Distance and its Application in Outlier Detection
基于距离的样本选择方法及其在实时野点检测中的应用

The paper introduces the approach of outlier detection based on kernel ,and points out the approach is hard to be realized with the increase of sample number. In order to reduce the size of optimization, the sample selection method based on distance is proposed. Through selecting samples,the computation workload and the demand of EMS memory can be decreased greatly. In the end,the real-time demand can be met.