多元线性回归是用多个自变量的最优组合共同来预测或估计因变量的一种方法。它具有模型简单,预测准确,解释能力强等特点,因而得到了广泛的应用。本文从用户研究中的一个实例入手,运用多元线性回归的方法建立预测模型,并对建模的每一步做了说明。在实际工作中,运用该预测模型获得了比较准确的结果。 Multivariate linear regression is a method of jointly predicting or estimating dependent variables using the optimal combination of multiple independent variables. It has a simple model, accurate prediction, strong ability to explain, and thus has been widely used. This paper starts from an example of user research, uses multiple linear regression methods to establish the forecasting model, and explains each step of modeling. In practical work, using the forecasting model obtained more accurate results.