2021-06-27 20:39 In statistics, a regression model is a mathematical model for quantitatively describing statistical relationships. Among them, linear regression is a fixed-effect statistical analysis method that determines the quantitative relationship between two or more variables. However, in actual statistical cases, the independent variables are often random, and we are not able to implement a well-designed matrix. A regression model in which the independent variable is a random variable is called a random effect model (variance component model). It is a layered linear model, and the analyzed data is extracted from different hierarchies, and the differences are related to the hierarchical structure.