Research Article
A Soft Computing Based Approach Using Modified Selection Strategy for Feature Reduction of Medical Systems
Table 3
Average time interval and memory usage levels of tested system softwares.
| |
Tested System | Number of Inputs | Time (average) | Allocated Memory Peak Working Set (MB), (average res.) |
| | | Modified Artificial Selection | 12 | 2–20 min. | 70–250 MB | 1 | FRSGR | Artificial Selection | 12 | 3.5–33 min. | 65–300 MB | Roulette Wheel Selection | 12 | 4–35 min. | 75–320 MB | | | Modified Artificial Selection | 15 | 4–30 min. | 75–320 MB |
| | |
12 |
70 min. |
380 MB | 2 |
Decision Relative Discenibility | | | | | | 15 | Exceeds 4 Hours | Exceeds 860 MB and causes memory error (insufficient memory) |
| 3 | Attribute dependency reduction without genetic search | 12 | Exceeds 2 Hours | Exceeds 899 MB and causes memory error |
|
|