Targeted stochastic gradient Markov chain Monte Carlo for hidden Markov models with rare latent states

no code implementations • 31 Oct 2018 • Rihui Ou, Deborshee Sen, Alexander L Young, David B Dunson

We propose a targeted sub-sampling (TASS) approach that over-samples observations corresponding to rare latent states when calculating the stochastic gradient of parameters associated with them.

Clustering Time Series +1