Abbreviation | Emission | Transition | Evaluated |
|---|
| | probabilities | probabilities | |
|---|
jpHMM prob | Probabilistic model | Joint |
✓
|
jpHMM ml | Machine learning | Joint |
✓
|
jpHMM scal | Pseudocount scaling | Joint |
✓
|
jpHMM semi | Original | Joint |
✓
|
jpHMM orig | Original | Separate | |
- The first column gives the abbreviation used for the respective model in what follows. The second column indicates how the emission probabilities are modeled: the probabilistic model introduced in this paper, the machine learning approach presented in [29], a scaling the pseudocounts depending on subtype size evaluated in [29], or the original algorithm introduced in [4]. The third column shows whether the transition probabilities are modeled jointly or separately. The last column indicates whether the respective approach was evaluated in this paper.
