Table 1 Comparison of cluster alignments, using a pipeline with a standard scaler, a dimension reduction method and a clustering model
From: Haematology dimension reduction, a large scale application to regular care haematology data
Dimension reduction | Components | # of clusters | NMI | ARI |
|---|---|---|---|---|
n/a | n/a | 12 | n/a | n/a |
PCA | 3 | 32 | 0.06 | 0.06 |
UMAP | 3 | 115 | 0.04 | 0.03 |
TriMap | 3 | 45 | 0.07 | 0.05 |
PaCMAP | 3 | 42 | 0.10 | 0.11 |
GRP | 3 | 30 | 0.01 | −0.0003 |
PCA | 6 | 31 | 0.04 | 0.03 |
UMAP | 6 | 84 | 0.09 | 0.10 |
TriMap | 6 | 44 | 0.05 | 0.01 |
PaCMAP | 6 | 43 | 0.05 | 0.02 |
GRP | 6 | 22 | 0.03 | 0.02 |
PCA | 12 | 12 | 0.07 | 0.09 |
UMAP | 12 | 81 | 0.08 | 0.09 |
TriMap | 12 | 53 | 0.08 | 0.08 |
PaCMAP | 12 | 54 | 0.06 | 0.05 |
GRP | 12 | 5 | 0.12 | 0.19 |