The colon
dataset of Spellman et. al. (1998)
Figure 1: Relative and cumulative variance of the yeast dataset
1. OKM
Figure 2: BIC values when applying OKM (SVD
reduced to 4 dimensions) on the yeast dataset
Figure 3: (zoom in) BIC Values when applying OKM (SVD reduced to 4
dimensions) on the yeast dataset
Figure 4: Comparison of
the internal (BIC) and external (Jaccard) criteria of the yeast dataset (OKM)
Figure 5: (zoom in) Comparison
of the internal (BIC) and external (Jaccard) criteria of the yeast dataset
(OKM)
2. OQM
Figure 6: Comparison of
the internal (BIC) and external (Jaccard) criteria of the yeast dataset (OQC)
|
|
Method |
Jaccard |
Purity |
Efficiency |
|
Raw
data |
|
|
|
|
|
|
K
Means (5 clusters) |
0.435 |
0.617 |
0.596 |
|
|
K
Means (4 clusters) |
0.488 |
0.64 |
0.673 |
|
|
Fuzzy
C Means (5 clusters) |
0.425 |
0.663 |
0.542 |
|
|
Fuzzy
C Means (4 clusters) |
0.438 |
0.458 |
0.912 |
|
|
Competitive
Neural Network (4 clusters) |
0.424 |
0.53 |
0.68 |
|
|
Quantum
Clustering |
NA |
NA |
NA |
|
Preprocessing |
|
|
|
|
|
|
K means (5 clusters) |
0.406 |
0.636 |
0.528 |
|
|
K means (4 clusters) |
0.46 |
0.626 |
0.634 |
|
|
Fuzzy C means (5 clusters) |
0.4 |
0.63 |
0.522 |
|
|
Fuzzy C means (4 clusters) |
0.459 |
0.624 |
0.634 |
|
|
Competitive Neural Network (5 clusters) |
0.33 |
0.55 |
0.458 |
|
|
Competitive Neural Network (4 clusters) |
0.516 |
0.658 |
0.706 |
|
|
QC after SVD – 2dims (σ =0.595) |
0.554 |
0.664 |
0.77 |
|
|
QC after SVD – 4 dims (σ =0.5) |
0.5 |
XXX |
XXX |
Table 1:
Comparison for algorithms’ performance the for the yeast dataset
