Skip to main content
Account

Table 3 Results (euros, 2012) for the base model and for ten single-constraint models

From: Improving risk equalization with constrained regression

 

R-squared (×100%)

CPM (×100%)

Base model

22.5%

24.8%

 Base model + single constraint to limit undercompensation for users of home care in t−1 by:

  

  20%

22.5%

24.9%

  40%

22.5%

24.9%

  60%

22.4%

24.9%

  80%

22.3%

24.9%

  100%

22.2%

24.8%

 Base model + single constraint to limit undercompensation for users of physiotherapy in t−1 by:

  

  20%

22.5%

24.9%

  40%

22.5%

25.0%

  60%

22.4%

25.0%

  80%

22.2%

25.0%

  100%

22.0%

24.8%