2018 Hiver Q3
8.
(a).
Write down the log-likelihood objective.
(b).
Show that maximizing this likelihood objective is equivalent to minimizing the KL divergence to the sampled data, DKL(pˆdata(x) ∥ pmodel(x; θ)).
Write down the log-likelihood objective.
Show that maximizing this likelihood objective is equivalent to minimizing the KL divergence to the sampled data, DKL(pˆdata(x) ∥ pmodel(x; θ)).