Skip to main content

Table 5 Average rank of autoencoder architectures over all small and large LFR networks, for must-link and cannot-link constraints

From: Overlapping community finding with noisy pairwise constraints

Architecture

Small networks

Large networks

Must-link constraints

Cannot-link constraints

Must-link constraints

Cannot-link constraints

AE1

3.0

4.5

2.5

4.0

AE1_L1

1.0

6.0

2.5

5.0

AE2

5.0

3.0

4.0

2.0

AE2_L1

2.0

4.5

1.0

6.0

AE3

6.0

2.0

6.0

3.0

AE3_L1

4.0

1.0

5.0

1.0