|
I could show you how EM
improves accuracy of the learned models, but for reasons of time, I will
focus on a more interesting use of the algorithm. Note that our algorithm learned from data
where Intelligence and Difficulty were never observed. It was essentially inventing these
concepts, i.e., defining for itself a concept of “smart” versus “weak”
students, based on the data. Of
course, in this example, the concepts were there to begin with, in the data
we generated. Does this work in real
data?
|