I need a clarification about the Cross validation in parametric model such as a simple linear regression: Are the coefficients of the final model the mean of the coefficients estimated in each round? Only in this case a can see a reduction in variance of the model, infact
$var(\overline{\beta_i}) = \sigma^2_{\beta_i}/n$
Different is for non parametric models such as decision trees, which they have hyper parameters, in that case is just model selection. Am I interpreting correctly?


Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Browse other questions tagged or ask your own question.