Creates a vector of conditional probabilities which is uded in most of the functions in this package.
Arguments
- HM
A HMM class object
- X
Data
- log_sum
True or False, If True then the log sum trick is used to deal with underflow.
Value
beta_matrix A matrix of backward probabilities
Details
Creates through a backward pass through the data a evaluation in in form of conditional probabilities
of the observations being x_t+1,...,x_t given that the Markov chain is in state i at time t.
