I want to use the #polr function of the #MASS package. What will be the factor of increase in computation time between a normal linear regression & f.e. a polr computation with the same input variables and an output with 15 levels? (Or even an output with 200 levels?)
Or are there better ways to model something with that much output levels?
#rstats #ordinalregression
#linearregression #logisticregression
Or are there better ways to model something with that much output levels?
#rstats #ordinalregression
#linearregression #logisticregression