Hello I am trying teach myself Logistical regression.
I am not quite sure how I should prepare my data for this function.
I work in maintenance for a fleet of vehicles.
I am trying to predict if a particular planned task will have unplanned tasks. (The mechanics find an issue with something while working the planned work, and create a separate task to work) Unplanned work costs us time and money, and we usually need to be aware of parts to order. This is definitely something we want to predict accurately.
So in order to do the calculation correctly, what kind of data should I feed the Logistic Regression formula?
Should miles/run time be a factor, if it is, should I categorize this as low = 1, mid-life = 2, high use = 3
or should I put the miles/hours operated into the formula as is?
should incidents like collisions within past year be a full count of all incidents? such as, if a vehicle had 5 incidents in the last year should it be counted as all 5 or simply Yes=1 No=0
Should "days since this task was last done" be factor? if so should I categorize the days into small numbers or allow the numbers be the exact number of days since a task last performed
I don't know enough about what the formula needs to properly function as a prediction of whether or not a Task will have findings for more work to be done
Any help is appreciated, thank you!
