ccf
ccf copied to clipboard
Predict probabilities
As far as I understand this implementation of CCF do not have a capability to predict probability of a class.
It would be nice to implement this.
There are still many things left open as to-do items. THese even hamper the basic functionality of the package at the moment. I hope to resolve it during the course of the year.
Best, Stefan
Am 22.06.2017 um 14:28 schrieb Oleg Kachan [email protected]:
As far as I understand this implementation of CCF do not have a capability to predict probability of a class.
It would be nice to implement this.
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or mute the thread.
This is true but can easily be fixed. The forrest prediction is done here
treePredictions <- apply(treePredictions, 1, function(row) {
names(which.max(table(row)))
})
Each row
contains the predictions of the individual trees. Instead of returning the class with maximal votes, this can be changed to compute the probabilities.