摘要： 多元线性回归被广泛用于预测。回归式反映了响应变量和预测变量间的线性关系。将模糊集理论引入多元线性回归中，通过模糊控制变量，可以得出更符合实际，也更容易为人所理解的回归模型。针对真实数据的实验表明，具有模糊控制变量的线性回归可以解决一类复杂的回归问题。

关键词: 多元线性回归；模糊控制变量；模糊集

Abstract: Multiple linear regression is used extensively in prediction. Regression expression reflects the linear relation between responding variable and predictive variable. By introducing fuzzy control variable, the paper proposes a fuzzy multiple linear regression model which can get more effective and more understandable fuzzy regression expression. Experiment on real data proves the efficiency of the fuzzy regression model.

Key words: Multiple linear regression; Fuzzy control variable; Fuzzy set