Table 28 Summary of part D scenarios, and optimal levels of features selected from each scenario
From: Data reduction techniques for highly imbalanced medicare Big Data
Scenario | Description |
|---|---|
Scenario One | RUS only: RUS 1:81 selected |
Scenario Two | Feature selection only: 10 features selected |
Scenario Three | Feature selection, then RUS: 10 features and RUS 1:81 selected |
Scenario Four | RUS, then feature selection: RUS 1:81 and 10 features selected |
Scenario Five | No preprocessing done on the data |