Implementation of 3 minimization algoritms proposed by prof. Yurii Nesterov in article Gradient methods for minimizing composite functions for function $ \Psi(x) $ in a form: $$ \Psi(x)=||b-Ax||_2 +\lambda||x||_1 $$ These algorithms can be used to create linear regression model with a built-in feature selection similar to Lasso.
- Adding stopping criterion from the paper
- Code optimisation
- adding decriptions to the functions
- making detailed README