SSM: Bayesian learning and inference for state space models
This package has fast and flexible code for simulating, learning, and performing inference in a variety of state space models. Currently, it supports:
Hidden Markov Models (HMM)
Auto-regressive HMMs (ARHMM)
Input-output HMMs (IOHMM)
Hidden Semi-Markov Models (HSMM)
Linear Dynamical Systems (LDS)
Switching Linear Dynamical Systems (SLDS)
Recurrent SLDS (rSLDS)
Hierarchical extensions of the above
Partial observations and missing data
We support the following observation models:
Gaussian
Student's t
Bernoulli
Poisson
Categorical
Von Mises
HMM inference is done with either expectation maximization (EM) or stochastic gradient descent (SGD). For SLDS, we use stochastic variational inferenc