Table 1 Hybrid algorithm construction of different sample sizes Car diagnosis Bayesian network modeling effect comparison
Sample size | algorithm | R(E) | M(E) | A(E) | S(E) |
|---|---|---|---|---|---|
100 | MMHC | 3.2 | 14.2 | 0.2 | 16.0 |
MMPC-Tabu | 3.2 | 14.2 | 0.2 | 16.0 | |
Fast.iamb-tabu | 2.6 | 16.4 | 0.2 | 17.9 | |
Inter.iamb-tabu | 3.2 | 14.2 | 0.2 | 16.0 | |
500 | MMHC | 4.2 | 11.7 | 0 | 13.8 |
MMPC-Tabu | 4.2 | 11.7 | 0 | 13.8 | |
Fast.iamb-tabu | 3.3 | 12.5 | 0 | 14.15 | |
Inter.iamb-tabu | 4.6 | 11.3 | 0 | 13.6 | |
1000 | MMHC | 4.6 | 10.5 | 0 | 12.8 |
MMPC-Tabu | 4.6 | 10.5 | 0 | 12.8 | |
Fast.iamb-tabu | 3.6 | 11.5 | 0 | 13.3 | |
Inter.iamb-tabu | 5.1 | 10.0 | 0 | 12.55 | |
5000 | MMHC | 5.2 | 9.8 | 0 | 12.2 |
MMPC-Tabu | 4.7 | 9.6 | 0.1 | 12.05 | |
Fast.iamb-tabu | 5.2 | 8.8 | 0.1 | 11.5 | |
Inter.iamb-tabu | 6.1 | 7.4 | 0 | 10.45 | |
10,000 | MMHC | 7 | 8.2 | 0 | 11.7 |
MMPC-Tabu | 5.7 | 8.1 | 0 | 10.95 | |
Fast.iamb-tabu | 5.3 | 6.9 | 0 | 9.55 | |
Inter.iamb-tabu | 5.4 | 5.4 | 0 | 8.1 | |
20,000 | MMHC | 7.9 | 7.4 | 0 | 11.35 |
MMPC-Tabu | 6.7 | 7.1 | 0 | 10.45 | |
Fast.iamb-tabu | 5.5 | 7.5 | 0 | 10.25 | |
Inter.iamb-tabu | 5.0 | 4.4 | 0 | 6.9 |