用Weighted fusion方法对森林中蝙蝠活动数据做变量选取

Variable selection for forest bats activities data by weighted fusion method

利用Weighted fusion变量选择方法,分析了森林中蝙蝠活动数据.计算了各变量之间的样本相关系数,画出了响应变量与预测变量之间的散点图,对预测变量按照重要性进行了排序,并且根据AIC与BIC准则选出了跟蝙蝠活动显著相关的预测变量.

This article analyzes forest bats activities data by weighted fusion variable selection method. First, it computes the sample correlation coefficient and obtains the scatter plot of the response variable and the predictor variables. Second, the order of predictors is given according to their importance and by weighted fusion method. Last,this article uses AIC and BIC to select the significant predictors related to forest bats activities.