Research Article
Predicting Protein-Protein Interactions via Random Ferns with Evolutionary Matrix Representation
Table 6
The prediction ability of the different methods on the H. pylori PPI dataset.
| | Related work | Method | ACC (%) | SN (%) | PE (%) | MCC (%) |
| | Martin et al.’s work [32] | | 83.40 | 79.90 | 85.70 | N/A | | You et al.’s work [53] | | 84.91 | 83.24 | 86.12 | 74.40 | | Nanni’s work [56] | WSR | 83.70 | 79.00 | 87.00 | N/A | | Bock and Gough’s work [57] | Phylogenetic Booststrap | 75.80 | 69.80 | 80.20 | N/A | | Nanni’s work [56] | LDC | 83.00 | 80.60 | 85.10 | N/A | | Shi et al.’s work [58] | Boosting | 79.52 | 80.37 | 81.69 | 70.64 | | Proposed method | MatFLDA_RFs | 85.35 | 95.72 | 79.27 | 74.41 |
|
|
Note: N/A means not available.
|