核工程中的数据挖掘

Data Mining in Nuclear Engineering

数据挖掘是在大量数据中发现有用的、人们感兴趣的信息的过程。支持向量机(SVM)是数据挖掘中的一项新技术;而支持向量回归机(SVR)是SVM在回归估计中的应用的体现。与传统的方法相比,SVR具有无事先人为强加性,直接由数据内在关系拟合而成,得到的结果更准确。本文介绍了SVR数学原理,并且利用SVR来处理核工程严重事故实验中熔融液滴运动的数据。

Data mining （DM） is a process to find the useful and interesting information in huge data. Support Vector Machine （SVM） is a new technique in data mining, but Support Vector Regression （SVR） is the applying of SM in regression . Compared with the traditional regression methods, SVR has not been specified beforehand, and is fitted directly from the inner relationship of data, thus the simulation results are more accurate. This paper introduces the mathematical theory of SVR and uses SVR to process the data of the moving characteristics of molten metal droplets in serious nuclear engineering accidents.