lx10077 / lasso

Convex optimizers for LASSO, including subgradient, project gradient, proximal gradient, smooth method, lagrangian method and stochastic gradient descent variants.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

The problem

Lasso (least absolute shrinkage and selection operator; also Lasso or LASSO) is a a regression analysis method that performs both variable selection and regularization in order to enhance the prediction accuracy and interpretability of the statistical model it produces. This repository contains various convex optimizers for it, including subgradient methods, project gradient, proximal gradient methods, smooth method, lagrangian methods and stochastic gradient descent variants.

Mathematically speaking, the problem is formed as $$\begin{equation} min_{x} \ \frac{1}{2}| Ax-b|^2 + \mu |x|_1 \end{equation}$$ where $A \in \mathbb{R}^{m \times n}$ and $b \in \mathbb{R}^{n}$ are given data matrix.

The Results

Common Convex Optimizers: Stochastic Gradient Descent Variants:

About

Convex optimizers for LASSO, including subgradient, project gradient, proximal gradient, smooth method, lagrangian method and stochastic gradient descent variants.


Languages

Language:Python 100.0%