What is the difference between soft and hard expectation maximization?
EDIT: ok, i've found out this paper: http://ttic.uchicago.edu/~dmcallester/ttic101-07/lectures/em/em.pdf that explain quite well the situations
Cross Validated is a question and answer site for people interested in statistics, machine learning, data analysis, data mining, and data visualization. It only takes a minute to sign up.
Sign up to join this communityWhat is the difference between soft and hard expectation maximization?
EDIT: ok, i've found out this paper: http://ttic.uchicago.edu/~dmcallester/ttic101-07/lectures/em/em.pdf that explain quite well the situations
Based on the paper you linked to, I would argue that the term EM usually refers to the "soft" version. The key distinction seems to be that instead of taking an expectation in the E-step, the "hard" version finds a mode.
A good explanation of the distinction is available in chapters 20-22 of David Mackay's book (which is available online).
I suggest you look at this: http://cogcomp.cs.illinois.edu/papers/SamdaniChRo12.pdf
There is a nice comparison in table 1.