Table 4 The three most important features of classification models
Model/Feature | Random Forest | Decision Tree | Gradient Boosting |
|---|---|---|---|
Most important feature | ControllableBlinds (0.210) | ControllableBlinds (0.240) | DifBetweenCurrentAnd IntendedTemp (0.175) |
Second most important feature | AirQualitySensor (0.180) | DifBetweenCurrentAnd IntendedTemp (0.180) | AirQualitySensor (0.120) |
Third most important feature | DifBetweenCurrentAnd IntendedTemp (0.95) | AirQualitySensor(0.90) | ControllableBlinds (0.120) |