I have some data where one of the columns is only relevant after evaluating a different column. For example, when analyzing car data to determine car values, it may be meaningful to evaluate whether a Toyota Camry is a Camry LE , and SE, or and XLE. But that is only meaningful after we determine it's a Camry first. If its a Honda, I know that those edition types are irrelevant, or perhaps even worse, counter predictive. Is there a way for me to help my RF model along by instructing it that certain decisions should always be made before sampling other columns?
I am pretty new to R, so it's taking me some time to reproduce my data. IN Addition, I was advised my mr flick that this should really be posted on the data science stack exchange. Thanks to all.