An Open Access Journal
Table 5 Importance estimates of predictor variables
From: Predicting the duration of motorway incidents using machine learning
Features | Regression trees | Classification trees |
|---|---|---|
Incident category | 0.0185 | 0.0370 |
Incident type | 0.4606 | 0.5062 |
Time period | 1.0 | 1.0 |
Day type | 0.0034 | 0.4431 |
Closed lane count | 0 | 0.1729 |
Precipitation | 0.2623 | 0.3557 |
Surf temperature | 0.2501 | 0.6191 |
Police attendance | 0.0241 | 0.1088 |
Fire service | 0.2288 | 0.1876 |
Ambulance | 0 | 0 |
Junction number | 0.0921 | 0.3771 |
Location to junction | 0.0334 | 0.0347 |
Distance to junction | 0.0594 | 0.2246 |