Dear coleagues,
Can someone help me on how to evaluate the output of the algorithm for blindfolding cross-validation method in SmartPLS from the values of CV and CV With Red, SSO and SSE estimates?
I also thank you any indication of references for supporting the oriented procedure.
Best Regards,
Simone
Blindfolding results interpretation
Hi.
Menu - Calculate - Blindfolding
In Blindfolding setting, insert Omission Distance(usually used from 5 to 10), and mark in endogenous constructs what you want to analyze.
Menu - Report - Default Report, and check Construct cross validated redundancy under results of Blindfolding where you can find SSO and SSE estimates that you want to find.
In addition, 1 - SSE/SSO is the predictive relevance, Q2, which is >0 is predictive relevance.
Menu - Calculate - Blindfolding
In Blindfolding setting, insert Omission Distance(usually used from 5 to 10), and mark in endogenous constructs what you want to analyze.
Menu - Report - Default Report, and check Construct cross validated redundancy under results of Blindfolding where you can find SSO and SSE estimates that you want to find.
In addition, 1 - SSE/SSO is the predictive relevance, Q2, which is >0 is predictive relevance.