Skip to main content

Table 2 Active learning performance results for all six data sets over 11 iterations of learning using the labeled set obtained from co-training. Results demonstrate average area under the precision-recall curves. Standard deviations were in the range of 0.001 to 0.003 for all experiments and are omitted from the table for clarity purposes

From: A semi-supervised machine learning framework for microRNA classification

Iteration

hsa

mmu

dme

bta

gga

eca

–

0.779

0.955

0.921

0.865

0.964

0.961

1

0.812

0.951

0.918

0.877

0.960

0.962

2

0.856

0.959

0.921

0.890

0.963

0.965

3

0.875

0.963

0.925

0.894

0.963

0.965

4

0.881

0.963

0.928

0.916

0.964

0.968

5

0.888

0.968

0.930

0.932

0.965

0.970

6

0.891

0.970

0.929

0.939

0.965

0.971

7

0.891

0.972

0.931

0.939

0.965

0.971

8

0.896

0.972

0.937

0.941

0.964

0.971

9

0.898

0.972

0.940

0.948

0.964

0.970

10

0.901

0.972

0.941

0.947

0.965

0.971

11

0.903

0.972

0.941

0.948

0.965

0.971