Hello everyone,
This is an odd situation. I want to use regression (ridge regression) for a data set which have many predictors but just have few observations. I know that I should eliminate some predictors to reduce the model complexity and deal with negative degree of freedom. However, if I want to use all predictors, can I clone the database to have more observation. For example on the image I attached, we have 6 predictors, but have only 2 observations (this is not a good example, I just want to use it to explain my idea). Can I clone the 2 observation to 12 observations and use them for the analysis? Can I do that? I am looking for any research method similar to this (i don't know what does it call).
Thank you all for your consideration,
- Badges
- Science topic
- Similar topics
- Computer Science
- Images