Skip to main content

Table 1 Performance comparison of PALMER and competing algorithms, including PALMER without constraints (“Unconstrained”), block EM (“BEM”), block classification EM (“BCEM”), block stochastic EM algorithms (“BSEM”), and block Gibbs sampler (“BGibbs”)

From: PALMER: improving pathway annotation based on the biomedical literature mining with a constrained latent block model

  Strong SNR Weak SNR
  Gene GO term Gene GO term
PALMER 0.00 (0.00) 0.01 (0.01) 0.05 (0.03) 0.02 (0.02)
Unconstrained 0.01 (0.05) 0.01 (0.03) 0.42 (0.11) 0.31 (0.11)
BEM 0.13 (0.21) 0.10 (0.15) 0.45 (0.04) 0.34 (0.06)
BCEM 0.10 (0.19) 0.08 (0.14) 0.45 (005) 0.36 (0.05)
BSEM 0.03 (0.11) 0.03 (0.09) 0.45 (0.08) 0.54 (0.08)
BGibbs 0.33 (0.18) 0.46 (0.20) 0.47 (0.03) 0.46 (0.05)
  1. Average and SD (within parenthesis) of error rates calculated over 100 simulated datasets are reported