@amol82 wrote:
This question is in reference to the below article
Would be great, if someone can help clarify. If I use c5.0, why doesnt the performance improve ?
data.rose <- ROSE(cls ~ ., data = hacide.train, seed = 1)data str(data.rose) 'data.frame': 1000 obs. of 3 variables: cls: Factor w/ 2 levels “0”,“1”: 1 1 1 1 1 1 1 1 1 1 …
x1 : num -0.618 -0.473 -0.168 0.567 0.52 ... x2 : num -0.8349 -1.1749 0.5181 -0.0321 -0.1633 …
prop.table(table(data.rose$cls))
0 1
0.52 0.48
newtree <- C5.0(x=data.rose[,2:3],y=data.rose$cls)
newpred <- predict.C5.0(newtree,hacide.test[,2:3])
accuracy.meas(hacide.test$cls,newpred)Call:
accuracy.meas(response = hacide.test$cls, predicted = newpred)
Examples are labelled as positive when predicted is greater than 0.5
precision: 0.020
recall: 1.000
F: 0.020
Posts: 1
Participants: 1