回归分析法在情报预测中的使用——基于“白宫钥匙”模型的选情分析  

The Use of Regression Analysis in Intelligence Forecasting——An Election Analysis Based on the Keys to the White House Model

作　　者：金龙 闻杰 JIN Long;WEN Jie(Institute of International Relations,National University of Defense Technology,Nanjing 210039)

机构地区：[1]国防科技大学国际关系学院

出　　处：《科技情报研究》2020年第2期59-66,共8页Scientific Information Research

基　　金：2018年度国家社会科学基金重大项目“海上丝绸之路地缘安全及风险管控研究”（编号：18ZDA130）

摘　　要：[目的/意义]随着情报分析手段不断演进,回归分析在情报预测中得到较多应用,文章试将回归分析运用于大选预测,通过选取适当变量来判断候选人得票走势,这对拓展情报预测方法的使用具有现实意义。[方法/过程]对“白宫钥匙”模型的关键要素进行判断,并在此基础上选取变量进行回归分析,以预测美国2020大选特朗普的得票情况。[结果/结论]在当前形势下,13个关键要素中现任政府的对立要素仅为4个,判断特朗普可以赢得2020年大选。普选得票率与肯定的关键要素数呈近似线性相关,应用分析后显示,特朗普的普选得票能超过半数,但在其剩下的任期内,仍存在诸多要素会改变两党的竞选态势。[Purpose/significance]With the continuous evolution of intelligence analysis methods,the use of regression analysis for intelligence prediction has been applied to some extent.This paper tries to apply regression analysis to general election prediction and judge the trend of candidates'votes by selecting appropriate variables for regression analysis,which is of practical significance for expanding intelligence analysis methods.[Method/process]By judging the key elements o‘f keys to the White House’model,variables were selected for regression analysis to predict the vote of Trump in the 2020 election.[Result/conclusion]In the current situation,only four of the 13 key elements are opposed to the current administration,judging that Trump can win the election in 2020.The percentage of the popular vote is nearly linearly correlated with the number of true key factors,and the regression analysis shows that Trump can get more than half of the popular vote,but there are still many factors that will change the campaign dynamics of both parties in the rest of his term.

关 键 词：回归分析 “白宫钥匙”模型 2020年大选 特朗普 情报预测 

分 类 号：G250.2[文化科学—图书馆学]