距离和
2基于DS的孤立点检测算法 针对孤立点检测算法的一些弊端,采用一个基于距离和(distance sum,DS)的孤立点检测算法。
提出了一种基于主分量分析和属性距离和的孤立点检测算法。
Then, via building a Measure function to represent gap between forecast-rainfall and actual-rainfall. Eventually, to take minimum of the sum of distance as evaluation function.而后,创建了一种距离函数来表征预测与实际降雨量之间的差距,最后用距离和的最小作为评价函数。
By making an analysis of different relations of distance, method of gravity center, method of mean value of class and method of sum of squares of deviations are compared in this paper.本文从各种距离之间的关系来对重心法、类平均法及离差平方和法作一些分析比较。