Skip to main content

Table 5 The result of clustering performance in pairwise relations, based on CCM of geodesic proximity \(G(v_i,v_j)\), temporal proximity \(P(v_i,v_j)\), and spatio-temporal proximity \(S(v_i,v_j)\) for three different linkage methods: complete, average and average group

From: A novel metric to measure spatio-temporal proximity: a case study analyzing children’s social network in schoolyards

Datasets

Metric

Silhouette

Calinski–Harabasz

Daviesi–Bouldin

Complete

Average

Average group

Complete

Average

Average group

Complete

Average

Average group

PG-1

\(G(v_i,v_j)\)

0.400*

0.380

0.312

7.865

9.763*

6.397

0.722*

0.622

0.994

 

\(P(v_i,v_j)\)

0.397

0.397

0.397

9.166

9.166

9.166

0.734

0.734

0.734*

 

\(S(v_i,v_j)\)

0.367

0.367

0.313

7.363

7.363

6.369

0.814

0.814

1.170

PG-2

\(G(v_i,v_j)\)

0.223

0.241

0.212

5.879

5.507

5.324

1.454

0.893

0.977

 

\(P(v_i,v_j)\)

0.292

0.292

0.249

8.276*

8.276*

8.006

0.666

0.666

1.465

 

\(S(v_i,v_j)\)

0.307

0.308*

0.263

7.438

6.713

6.637

0.608*

0.733

1.583

  1. The best performance per linkage method is indicated in boldface, and the best performance among all linkage methods is shown with \((*)\)