Under high dimensional data, the number of parameters, p is greater than the sample size, n. REcently, many biomedical and social media are trending inside this problem, hence making it a popular statistical problem. Anyone with a solution providing a theoretical background?
See the paper attached below. It should be helpful, as will be this link:
https://www.google.com/search?q=p%3En+lasso&ie=utf-8&oe=utf-8&client=firefox-b-1-ab
Best wishes, David Booth
Dear Prof. Booth,
Thank you for the pointers to the solution for this problem. Do you have further pointers towards the theoretical derivatives of p>n problem? Why and how it becomes a statistical problem?
Ronald.
Yes. See the link below especially the paper by Trevor Hastie. Best wishes, David Booth
https://www.google.com/search?client=firefox-b-1&ei=c8dsW_2gNIfJjwT79oGYBA&q=+p%3En++problem+for+ordinary+regression&oq=+p%3En++problem+for+ordinary+regression&gs_l=psy-ab.3...9025.27663.0.30299.7.7.0.0.0.0.278.974.3j2j2.7.0....0...1c.1.64.psy-ab..0.0.0....0.Ani27hMFEYY
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size