| Issue | #Downvotes for this reason | By |
|---|
| C | Covariance Type: whether the covariances of gaussian outputs should be full matrices or limited to diagonal or spherical matrices | |
| D | Tied Covariance: whether the covariances of gaussian outputs are tied to be the same across all outputs | |
| I | Iteration Cutoff: the proportional minimum change of likelihood at which to stop the EM iteractions | |
| L | Left Right: whether the state transitions are constrained to go only to the next state in numerical order | |
| R | Random Initialisation: whether the state transition probabilities are intialized randomly (if this is false they are initialised by performing a k-means clustering on the data) | |
| S | States: number of HMM states to use |