Table 5 Comparison of LR, LASSO, and RF models performance on the test set in different scenarios, assessing models’ stability
From: A model to forecast the two-year variation of subjective wellbeing in the elderly population
Scenario | Metric | Method | ||
|---|---|---|---|---|
LR | LASSO | RF | ||
Main model | accuracy | 64% | 64% | 65% |
AUC | 0.692 | 0.694 | 0.659 | |
F1 score | 0.663 | 0.662 | 0.710 | |
Models’ stability | accuracy | 61% | 61% | 61% |
AUC | 0.610 | 0.614 | 0.608 | |
F1 score | 0.630 | 0.629 | 0.667 | |
Variables’ stability | accuracy | 63% | 63% | 62% |
AUC | 0.629 | 0.630 | 0..615 | |
F1 score | 0.649 | 0.645 | 0.659 | |