Table 2 Error analysis. Shown are the numbers of correctly predicted proteins, the numbers of correctly predicted footprints, and the number of times each method predicted the most probable of the footprints found by the two methods and the true footprint.
From: New decoding algorithms for Hidden Markov Models using distance measures on labellings
Algorithm | correct footprint | Phobius measure | most probable footprint | τ= 5 |
|---|---|---|---|---|
Phobius | 169 | 166 | 201 | 79 |
10 samples, r = 6 | 164 | 162 | 241 | 92 |