Linear regression assumes a linear relationship, is sensitive to outliers, and may not perform well if the assumptions (like homoscedasticity or normality) are violated.

Discover how linear regression works, from simple to multiple linear regression, with step-by-step examples, graphs and real-world applications.

An accountant can use linear regression only if he can apply the linearity assumption to the cost he is predicting.

Journal of Applied Econometrics, Vol. 24, No. 4 (Jun. - Jul., 2009), pp. 651-674 (24 pages) We consider the problem of variable selection in linear regression models. Bayesian model averaging has ...