Table 1.
Comparison of time to complete analysis of various clustering algorithms using the centered data
| Simulated dataset |
Experimental dataset |
|||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Genes | 10 000 |
10 000 |
22 283 |
44 760 |
||||||||
| Clusters |
4 |
10 |
20 |
4 |
10 |
20 |
4 |
10 |
20 |
4 |
10 |
20 |
| HPC-DEN | 0.39±0.04 | 0.42±0.03 | 0.46±0.03 | 0.32±0.01 | 0.36±0.02 | 0.67±0.08 | 0.49±0.07 | 0.84±0.07 | 1.48±0.14 | 1.01±0.17 | 2.16±0.25 | 3.37±0.74 |
| HPC-RIA | 0.40±0.05 | 0.62±0.06 | 0.92±0.09 | 0.33±0.02 | 0.41±0.02 | 0.71±0.09 | 0.58±0.10 | 0.92±0.13 | 1.58±0.15 | 1.27±0.14 | 2.41±0.30 | 4.22±0.44 |
| R-KM | 0.23±0.02 | 0.56±0.07 | 1.61±0.09 | 0.24±0.02 | 0.86±0.08 | 2.94±0.10 | 0.59±0.04 | 2.00±0.09 | 9.02±0.49 | 1.07±0.07 | 4.72±0.19 | 22.68±1.39 |
| R-SOM | 0.22±0.01 | 0.48±0.04 | 0.80±0.01 | 0.21±0.01 | 0.43±0.01 | 0.83±0.01 | 0.44±0.01 | 0.93±0.03 | 1.68±0.03 | 1.01±0.02 | 1.94±0.04 | 3.48±0.02 |
| Cluster v2 | 0.35±0.02 | 0.62±0.12 | 1.95±0.45 | 1.48±0.08 | 5.09±0.49 | 11.80±1.20 | 4.72±0.61 | 25.94±1.32 | 45.25±8.81 | 9.02±2.09 | 47.75±9.67 | 118.33±6.51 |
| Cluster v3 | 1.12±0.08 | 3.06±0.18 | 8.94±0.50 | 3.70±0.42 | 9.68±0.54 | 19.28±1.61 | 8.73±0.28 | 26.90±1.88 | 57.50±4.95 | 11.70±0.56 | 51.50±2.89 | 162.33±10.60 |
HPC-DEN, density from data; HPC-RIA, random initial assignment; R-KM, R-k-mean; R-SOM, R self organizing maps; Cluster v2, Cluster v3, k-means algorithm.
All values for the programs are the average time to completion in minutes and the standard deviation.
N = 12 for each. Bold items indicate significantly lower times within each column.