What is it about?
In statistics, a well-known model is the linear regression. With it, people learn the impact of feature variables on the output. Yet, in practice, people may tentatively have MANY feature variables to select. To select a FEW key feature variables, one can use Lasso. However, it is non-trivial to numerically solve Lasso. In this paper, we review the state-of-the-art algorithms to solve Lasso.
Featured Image
Photo by Clément Hélardot on Unsplash
Why is it important?
There is a lot of literature available, discussing the statistical properties of Lasso. However, there lacks a comprehensive review discussing the algorithms to solve Lasso. And this paper bridges this gap.
Perspectives
This is beginner-friendly reading material for people who are interested in statistics, data science, and optimization research. It starts with hands-on tutorials and ends with thoughtful explanations of complicated technical details.
Yujie Zhao
Merck and Co Inc
Read the Original
This page is a summary of: A survey of numerical algorithms that can solve the Lasso problems, Wiley Interdisciplinary Reviews Computational Statistics, October 2022, Wiley,
DOI: 10.1002/wics.1602.
You can read the full text:
Contributors
The following have contributed to this page
