hard20 pts · first answer countsmachine-learningdecision-treesforecasting

Forest forecast that goes flat

Monthly sales have risen steadily for three years. A random forest using "month number" as a feature predicts next year's sales as flat at about the last level. Why?

A.The forest needs many more trees before it sees the trendB.Random forests ignore numeric features such as monthC.Month number must be one-hot encoded for trees to use itD.Leaves output means of seen targets, so trees cannot extrapolate

Discussion 0

Join free to post in the discussion.

Forest forecast that goes flat

Discussion 0

Related questions