On the Adversarial Robustness of LASSO Based Feature Selection
On the Adversarial Robustness of LASSO Based Feature Selection
In this chapter, we investigate the adversarial robustness of LASSO based feature selection problem. We introduce a smooth approximation of the ℓ 1 norm and use the projected gradient descent to design the modifications on the feature matrix and response values in order to manipulate the regression coefficient. This chapter …