E-Step of the EM Algorithm
Holding the current parameters fixed, compute every responsibility gamma(z_nk); this builds the expected complete-data log-likelihood Q(theta, theta_old), replacing the unknown hard assignments with their posterior expectations.
This Concept is waiting for its first lesson!
Holding the current parameters fixed, compute every responsibility gamma(z_nk); this builds the expected complete-data log-likelihood Q(theta, theta_old), replacing the unknown hard assignments with their posterior expectations.
Are you a teacher? Sign in to start contributing.
Sign In