Recency-weighted Markovian inference

11/08/2017

∙

by   Kristjan Kalm, et al.

∙

0

∙

share

We describe a Markov latent state space (MLSS) model, where the latent state distribution is a decaying mixture over multiple past states. We present a simple sampling algorithm that allows to approximate such high-order MLSS with fixed time and memory costs.