Skip to main content

Table 6 Effects of training data size on prediction performance (AUC) for human

From: Protein-protein interaction prediction based on multiple kernels and partial network with linear programming

 

G tt ∼2600

G tt ∼2100

G tt ∼1600

\(\phantom {\dot {i}\!} {RL}_{ {WOLP-K-1}:G_{tn}\sim 3310} \)

0.9277

-

-

\( {RL}_{G_{tn}\sim 3710} \)

0.8359

-

-

\(\phantom {\dot {i}\!} {RL}_{ {EW-K}:G_{tn}\sim 3310} \)

0.8590

-

-

\(\phantom {\dot {i}\!} {RL}_{ {WOLP-K-1}:G_{tn}\sim 3310} \)

-

0.9305

-

\( {RL}_{G_{tn}\sim 4210} \)

-

0.8779

-

\(\phantom {\dot {i}\!} {RL}_{ {EW-K}:G_{tn}\sim 3310} \)

-

0.8620

-

\(\phantom {\dot {i}\!} {RL}_{ {WOLP-K-1}:G_{tn}\sim 3310} \)

-

-

0.9338

\( {RL}_{G_{tn}\sim 4710} \)

-

-

0.9227

\(\phantom {\dot {i}\!} {RL}_{ {EW-K}:G_{tn}\sim 3310} \)

-

-

0.8639