Table 14 The performance for mini-batch K-means algorithm for the second situation Preprocessing 2.
From: Open source Arabic research paper dataset for natural language processing
# K | Silhouette coefficient (euclidean) | Silhouette coefficient (cosine) | ARI | Davies–Bouldin Index |
|---|---|---|---|---|
2 | 0.029 | 0.053 | 0.121 | 5.716 |
3 | 0.0160 | 0.029 | 0.048 | 2.554 |
4 | 0.022 | 0.037 | 0.057 | 2.906 |
5 | 0.043 | 0.078 | 0.150 | 3.865 |
6 | 0.008 | 0.011 | 0.150 | 3.146 |
7 | 0.023 | 0.041 | 0.100 | 2.649 |
8 | -0.008 | -0.018 | 0.179 | 2.905 |
9 | -0.003 | -0.009 | 0.205 | 2.908 |
10 | 0.029 | 0.053 | 0.283 | 3.699 |
