Linear regression is a statistical method used for modeling the relationship between a dependent variable and one or more independent variables. It assumes a linear relationship and estimates the coefficients of the model to minimize the difference between predicted and actual values. This technique is widely used for forecasting and trend analysis in various fields.