Research Article

Sparse Graph Embedding Based on the Fuzzy Set for Image Classification

Table 5

The standard deviation, best average recognition rates (%), and corresponding feature dimension of various algorithms in the AR face database.

Methods
23456

PCA66.68 ± 1.11 (145)70.21 ± 1.62 (145)77.60 ± 1.27 (150)79.03 ± 0.81 (150)81.54 ± 1.04 (145)
LDA71.5 ± 0.71 (110)75.58 ± 0.76 (115)82.53 ± 0.81 (120)87.12 ± 0.33 (115)87.58 ± 0.75 (120)
MMC69.39 ± 1.15 (120)75.71 ± 0.59 (115)82.98 ± 0.62 (115)85.27 ± 0.94 (120)87.86 ± 0.82 (120)
LPP72.82 ± 0.81 (85)78.78 ± 1.10 (85)83.70 ± 0.65 (80)88.75 ± 0.31 (85)89.93 ± 0.48 (70)
LLE70.40 ± 0.89 (125)74.31 ± 1.16 (145)83.35 ± 1.49 (140)84.78 ± 0.76 (150)86.98 ± 0.40 (145)
SDE74.90 ± 0.08 (85)80.10 ± 1.36 (75)84.35 ± 1.36 (80)89.80 ± 0.65 (85)92.50 ± 1.18 (70)
FLGE/MMC75.09 ± 1.49 (75)82.42 ± 1.08 (70)88.23 ± 0.8 (75)90.17 ± 1.36 (75)93.38 ± 1.20 (70)
The proposed method79.31 ± 1.65 (80)85.79 ± 0.75 (90)90.86 ± 0.82 (95)92.74 ± 0.95 (85)95.82 ± 0.68 (90)