The LASSO (Least Absolute Shrinkage and Selection Operator) method, developed by Robert Tibshirani in 1996, efficiently predicts outcomes while maintaining an accurate and minimalist model. In LASSO regression, the objective function minimizes the residual sum of squares (RSS) plus a penalty term involving a regularization parameter (λ) and coefficients (β_j) for predictors.
The penalty term encourages coefficient shrinkage towards zero, balancing data fit and model simplicity.