Table 1.
Dataset | PCA-Km | LE-Km | PCA-DP | GMMsort | LDA-GMM | LDA-DP |
---|---|---|---|---|---|---|
C1_005 | 100 (0.0) | 100 (0.0) | 100 (0.0) | 100 (0.0) | 100 (0.0) | 100 (0.0) |
C1_010 | 100 (0.0) | 100 (0.0) | 100 (0.0) | 99.9 (0.2) | 100 (0.0) | 100 (0.0) |
C1_015 | 99.9 (0.1) | 99.8 (0.1) | 99.9 (0.1) | 99.9 (0.1) | 99.9 (0.1) | 99.9 (0.1) |
C1_020 | 99.4 (0.3) | 98.8 (0.6) | 99.2 (0.4) | 97.5 (2.3) | 92.9 (10.3) | 99.4 (0.3) |
C1_025 | 97.6 (0.2) | 96.7 (0.6) | 97.2 (0.3) | 96.5 (0.9) | 91.7 (12.6) | 97.7 (1.0) |
C1_030 | 93.8 (1.5) | 81.8 (22.9) | 81.7 (21.0) | 92.0 (2.9) | 97.5 (1.0) | 96.4 (2.3) |
C1_035 | 66.9 (22.7) | 78.6 (19.7) | 76.1 (11.9) | 76.7 (15.8) | 93.3 (2.3) | 94.9 (3.2) |
C1_040 | 62.5 (21.7) | 79.8 (13.6) | 67.3 (12.8) | 69.7 (12.4) | 83.3 (16.6) | 90.7 (8.5) |
C2_005 | 100 (0.0) | 100 (0.0) | 100 (0.0) | 100 (0.0) | 100 (0.0) | 100 (0.0) |
C2_010 | 98.2 (0.6) | 99.9 (0.1) | 98.2 (0.6) | 81.6 (16.7) | 86.8 (16.2) | 93.2 (13.5) |
C2_015 | 87.5 (1.2) | 96.4 (0.8) | 79.5 (9.6) | 66.3 (17.3) | 85.4 (18.0) | 86.3 (16.7) |
C2_020 | 72.9 (1.1) | 82.3 (1.6) | 36.8 (4.9) | 61.9 (5.2) | 96.5 (6.4) | 97.0 (5.5) |
C3_005 | 99.6 (0.3) | 100 (0.0) | 100 (0.0) | 93.4 (13.1) | 96.2 (7.5) | 100 (0.0) |
C3_010 | 89.9 (0.9) | 97.2 (0.9) | 73.9 (12.4) | 77.2 (14.2) | 95.6 (8.8) | 91.8 (11.7) |
C3_015 | 76.3 (2.1) | 85.0 (0.8) | 42.3 (9.7) | 80.9 (12.2) | 93.2 (12.7) | 99.6 (0.3) |
C3_020 | 55.7 (9.3) | 62.6 (6.9) | 37.2 (6.5) | 43.9 (8.5) | 70.3 (29.1) | 91.1 (11.0) |
C4_005 | 99.6 (0.4) | 100 (0.0) | 99.8 (0.1) | 100 (0.0) | 100 (0.0) | 100 (0.0) |
C4_010 | 94.1 (1.0) | 97.2 (0.7) | 73.1 (10.7) | 94.5 (7.7) | 96.5 (7.0) | 99.8 (0.3) |
C4_015 | 63.4 (19.6) | 68.2 (22.7) | 65.7 (4.5) | 68.7 (8.1) | 96.5 (5.9) | 96.9 (1.2) |
C4_020 | 50.9 (15.0) | 58.6 (4.9) | 43.5 (10.8) | 49.0 (15.0) | 90.3 (13.3) | 88.4 (4.9) |
Average | 85.4 (5.4) | 89.1 (5.4) | 78.5 (6.5) | 82.4 (8.5) | 93.3 (9.3) | 96.2 (4.5) |
The standard deviation of the accuracy is in parenthesis. The bold number represents the best performance in each testing set.