Linear dynamical systems (aka Kalman filters)
It provides matlab packages to learn parameters for linear dynamical systems (kalman filters). It contains the following package:
dynammo Expectation-Maximization algorithm to learn the parameters of a linear dynamical system. There also also additional code for learning with missing observations and recovering the missing values. located in the dynammo subdirectory check out make demo
PLiF a package for extracting shift invariant and harmonics features from time series.
Bolero a package for missing value imputation for human motion capture sequence, utilizing the human body skeleton
CLDS a package for learning complex-valued dynamical systems