The adaptive lasso is a modified version of the lasso that addresses its limitations in variable selection consistency. The lasso, a popular method for simultaneous estimation and variable selection, can be inconsistent under certain conditions. The adaptive lasso introduces adaptive weights in the $\ell_1$ penalty to achieve the oracle properties, meaning it performs as well as if the true model were known. It is also near-minimax optimal and can be solved efficiently using the same algorithm as the lasso. The adaptive lasso extends to generalized linear models and maintains the oracle properties under mild conditions. The nonnegative garotte is shown to be consistent for variable selection. The adaptive lasso is computationally efficient and performs well in simulations compared to other methods. It is particularly effective in high-dimensional settings and has been shown to outperform the lasso and SCAD in variable selection and prediction accuracy. The adaptive lasso's oracle properties do not automatically guarantee optimal prediction performance, but it is advantageous in difficult prediction problems. The paper concludes that the adaptive lasso is a valuable tool in statistical modeling and learning.The adaptive lasso is a modified version of the lasso that addresses its limitations in variable selection consistency. The lasso, a popular method for simultaneous estimation and variable selection, can be inconsistent under certain conditions. The adaptive lasso introduces adaptive weights in the $\ell_1$ penalty to achieve the oracle properties, meaning it performs as well as if the true model were known. It is also near-minimax optimal and can be solved efficiently using the same algorithm as the lasso. The adaptive lasso extends to generalized linear models and maintains the oracle properties under mild conditions. The nonnegative garotte is shown to be consistent for variable selection. The adaptive lasso is computationally efficient and performs well in simulations compared to other methods. It is particularly effective in high-dimensional settings and has been shown to outperform the lasso and SCAD in variable selection and prediction accuracy. The adaptive lasso's oracle properties do not automatically guarantee optimal prediction performance, but it is advantageous in difficult prediction problems. The paper concludes that the adaptive lasso is a valuable tool in statistical modeling and learning.