Figure 1
From: A binary matrix factorization algorithm for protein complex prediction
Prediction evaluations of BMF and MCL on altered graphs Prediction evaluations of BMF and MCL on altered graphs constructed from a test graph, with add% edges randomly added and/or del% edges randomly deleted with respect to the original number of edges. To save computation, we actually implement BMF(opt) at the same initialization as the one used in Table 1, and use the optimal inflation parameter values given by the Table (2) in [3].