Assumptions of linear regression

Linear regression assumes that the relationship between the variables can be described with a straight line. It presumes that this relationship is consistent across all data points, meaning the pattern of how one variable affects the other stays the same. The method also assumes that the data points are independent, with no pattern or trend that depends on each other. Additionally, it expects the errors (differences between observed and predicted values) to be randomly distributed, with no systematic pattern, and that these errors have similar variance throughout the data. These assumptions help ensure the model's predictions are accurate and reliable.