Table 2 |
||||
|
Results for four data sets |
||||
|
AR/GR/PR |
GATA |
NF-κB |
Thyroid |
|
|
|
||||
|
ML |
54.7 |
77.0 |
81.6 |
51.3 |
|
|
||||
|
MCL |
55.2 |
73.2 |
76.5 |
50.0 |
|
|
||||
|
MAP |
55.1 |
77.0 |
81.6 |
51.3 |
|
|
||||
|
MSP |
56.9 |
77.0 |
79.6 |
50.4 |
|
|
||||
|
Unified |
57.3 |
77.5 |
81.8 |
52.3 |
|
|
||||
|
Summary the results of Figure 2 for the 4 data sets containing the highest sensitivity for the ML, the MCL, the MAP, the MSP, and the unified generative-discriminative learning principle. For the MAP, the MSP, and the unified generative-discriminative learning principle, we present the best results form the simplex β which correspond to one of these learning principles (see Figure 1b). For each data set, the highest sensitivity is displayed in bold. |
||||
|
Keilwagen et al. BMC Bioinformatics 2010 11:98 doi:10.1186/1471-2105-11-98 |
||||
