Process:
- Select unique records (unique reference numbers)
- Calculate average of Number of Vehicles
(Result: 1.817008)- Calculate sum of accidents for each time bucket (group by time bucket)
(Result: Evening: 4,663)- Filter on Casualty Class = "Pedestrian"
- Count Distinct Reference Number
(Result: 2,562)- Create field WasFatal as boolean
- Oversample tool on field WasFatal = Yes at 50% level
- Logistic Regression tool on all the fields noted
(- Casualty Class with lowest P Value: Pedestrian: 0.00011- Coefficient estimate: 1.86759- Gender: Male)