Bayesian Estimation of Change-Points in the Slope of Multivariate Time-Series
tools:::Rd_package_title("beast")
Main function
Birth Probabilities
Complexity prior distribution
Compute the empirical mean.
Compute empirical posterior parameters
Posterior parameters
Move 3.b
Log-likelihood function.
Log-prior.
MCMC sampler
Printing
Zero normalization
Plot function
Print function
Move 2
Prior random numbers
Generate change-points according to the prior
Move 3.b
Truncated Poisson pdf
Move 1
Assume that a temporal process is composed of contiguous segments with differing slopes and replicated noise-corrupted time series measurements are observed. The unknown mean of the data generating process is modelled as a piecewise linear function of time with an unknown number of change-points. The package infers the joint posterior distribution of the number and position of change-points as well as the unknown mean parameters per time-series by MCMC sampling. A-priori, the proposed model uses an overfitting number of mean parameters but, conditionally on a set of change-points, only a subset of them influences the likelihood. An exponentially decreasing prior distribution on the number of change-points gives rise to a posterior distribution concentrating on sparse representations of the underlying sequence, but also available is the Poisson distribution. See Papastamoulis et al (2017) <arXiv:1709.06111> for a detailed presentation of the method.