From: Topic modeling for cluster analysis of large biological and medical datasets
Methods | k | Cluster ID | Adenocarcinoma | Squamous cell carcinoma | No. of misclassified samples | NMI |
---|---|---|---|---|---|---|
Clustering based on feature selection | 2 | 1 | 42 | 11 | 22 | 0.2809 |
 |  | 2 | 11 | 47 |  |  |
 | 3 | 1 | 40 | 8 | 21 | 0.2417 |
 |  | 2 | 4 | 15 |  |  |
 |  | 3 | 9 | 35 |  |  |
 | 4 | 1 | 37 | 8 | 18 | 0.2926 |
 |  | 2 | 9 | 35 |  |  |
 |  | 3 | 0 | 14 |  |  |
 |  | 4 | 7 | 1 |  |  |
Clustering based on highest topic assignment | 2 | 1 | 13 | 46 | 25 | 0.2296 |
 |  | 2 | 40 | 12 |  |  |
 | 3 | 1 | 11 | 29 | 25 | 0.1847 |
 |  | 2 | 37 | 9 |  |  |
 |  | 3 | 5 | 20 |  |  |
 | 4 | 1 | 5 | 13 | 26 | 0.1744 |
 |  | 2 | 13 | 26 |  |  |
 |  | 3 | 1 | 12 |  |  |
 |  | 4 | 34 | 7 |  |  |
Clustering based on feature extraction | 2 | 1 | 13 | 47 | 24 | 0.2461 |
 |  | 2 | 40 | 11 |  |  |
 | 3 | 1 | 8 | 34 | 24 | 0.2055 |
 |  | 2 | 8 | 16 |  |  |
 |  | 3 | 37 | 8 |  |  |
 | 4 | 1 | 7 | 6 | 25 | 0.1820 |
 |  | 2 | 33 | 6 |  |  |
 |  | 3 | 8 | 31 |  |  |
 |  | 4 | 5 | 15 |  |  |