作　　者： ;

机构地区： 汕头大学

出　　处： 《金融》 2019年第4期390-408,共19页

摘　　要： 本文主要关注潜在的人口因素与经济指标之间的关系,以期提供经济发展的建议。首先,对原始数据进行预处理,包括数据获取、截取、重命名、标准化等,使数据易于读取和识别。然后用相关分析来判断因素存在的正或负相关影响,以及哪个因素是显著的,哪个因素是不显著的。之后,使用线性回归、多项式回归、非线性回归三种方法来探索一个因素与经济指标之间的关系,找到更好的匹配方法。最后通过使用四种方法进行多因素分析,包括直接使用多元线性回归,主成分分析(PCA) + 多元线性回归,逐步回归,时间序列,找到一个最为合适的回归方法。 This paper mainly focuses on the relationship between the potential factors with economic indices in order to provide developing suggestions.First,we preprocess the original data,including data reading,deleting,rename,standardization and so on,to make the data easily read and recognized. Then,we use correlation analysis to tell which factor is positive and negative and the death is not a strong linear factor to economic indexes,while others are.After that,we use three methods to ex-plore the relationship between one factor and one of the economic indexes,including linear re-gression,polynomial regression as well as nonlinear regression,and we find out utilizing the re-sidual is the least to determine the better way to match the relationship.Finally,we use five me-thods to do the multiple factors analysis,including directly using multiple regression,PCA+mul-tiple regression,stepwise regression,time series analysis,to determine the best model for pre-dicting the economic development among them.

关 键 词： 经济指标 因素 人口 分析 方法 经济 预处理 线性 发展

领　　域： []