You are developing a regression model to forecast the next-quarter energy usage of a large manufacturing plant. The training set has 20 000 rows and roughly 400 engineered features from industrial sensors, many of which are highly correlated. An ordinary least-squares model overfits and shows high validation error. The stakeholders insist on a linear model that (1) applies coefficient shrinkage to reduce variance, (2) can drive some coefficients exactly to zero to eliminate redundant sensors, and (3) remains stable in the presence of strongly correlated predictors. Which regressor best satisfies all of these requirements?
Elastic Net regression simultaneously applies an L1 penalty (like LASSO) and an L2 penalty (like Ridge). The L1 term can set some coefficients to zero, performing automatic feature selection and removing redundant sensor channels, while the L2 term stabilizes coefficient estimates when predictors are highly correlated and reduces variance. LASSO alone performs feature selection but can behave unpredictably with multicollinearity; Ridge controls variance but never eliminates redundant predictors; a decision tree regressor is nonlinear and does not provide the desired linear coefficient interpretation. Therefore Elastic Net is the most appropriate choice for the stated requirements.
Ask Bash
Bash is our AI bot, trained to help you pass your exam. AI Generated Content may display inaccurate information, always double-check anything important.
Why is Elastic Net regression better than LASSO or Ridge for this scenario?
Open an interactive chat with Bash
What is multicollinearity and why is it a problem in regression models?
Open an interactive chat with Bash
How do L1 and L2 penalties improve regression models?