Table 6 Indicators of CS features with different dimensions on dataset A
From: Feature extraction method for proteins based on Markov tripeptide by compressive sensing
Feature vector | tr(S w )/ tr(S b ) | Clustering accuracy |
|---|---|---|
5 - dimensional CS features | 7.83197 | 0.8460 |
10 - dimensional CS features | 7.82934 | 0.8460 |
15 - dimensional CS features | 7.82953 | 0.8460 |
20 - dimensional CS features | 7.83036 | 0.8460 |
30 - dimensional CS features | 7.83051 | 0.8460 |
50 - dimensional CS features | 7.83031 | 0.8460 |