October 2017
Beginner to intermediate
572 pages
26h 1m
English
We have used Super Learner to fit the single model. We can specify more models also. We tried with random forest and glment with family as gaussian for a continuous outcome. For a binary outcome family is binomial. Risk shows model accuracy or performance--the lower the risk, the better the model. It shows that the model is predicting more accurate values. The coefficient column tells the weight of a model.
Read now
Unlock full access