I propose the idea of the best model having minimum mean of error rate in the validation sample by the k-fold cross validation for small sample method. See my papers.
Respected sir I gone through your paper but I could not identify where is written about model selection. Please help .. If you wish I can attach my file for more convenience
Sorry.
Papers related model selection are papers after 2013.
You check the papers of SOIC, Springer, Data Analytics.
After I return to my home next Thuesday, I decide to show the new one or not.
I define the best model having the minimum mean of error rates in the validation sample among seven LDFs.
You can do the fitting with several models and use the sum of squares error (SSE) as the criteria for choosing the best model.
Inimfon, I don't agree your answer.
I think the best model is the minimum mean of error rate in the validation sample among all comparison methods.
I am sorry, Inimfon.
If Mihir deails with regression model, we must use SSE instead minimum mean errror rate that is used in the discriminant analysis.
- Badges
- Science topic