Table 2 Average classification accuracy (± standard deviation) for each feature type
From: A machine learning approach to predict perceptual decisions: an insight into face pareidolia
Subject | Classification performance of individual subjects (in %) | |||
|---|---|---|---|---|
TFPS64 | TFPSL | TFPSR | DATFPS | |
p value: 0.025 | p value: 0.04 | p value: 0.025 | p value: 0.035 | |
Subject1 | \(\mathit{74 }.\mathit{80 } \pm \mathit{6 }.\mathit{06 }\) | \(69.60 \pm 6.41\) | \(67.18 \pm 6.94\) | \(73.33 \pm 6.76\) |
Subject2 | \(74.45 \pm 6.63\) | \(70.89 \pm 7.20\) | \(68.60 \pm 6.84\) | \(\mathit{77 }.\mathit{24 } \pm \mathit{7 }.\mathit{38 }\) |
Subject3 | \(68.12 \pm 5.84\) | \(65.53 \pm 5.59\) | \(65.01 \pm 6.07\) | \(\mathit{73 }.\mathit{17 } \pm \mathit{6 }.\mathit{95 }\) |
Subject4 | \(74.59 \pm 6.23\) | \(73.64 \pm 5.90\) | \(67.44 \pm 6.09\) | \(\mathit{77 }.\mathit{32 }\pm \mathit{6 }.\mathit{59 }\) |
Subject5 | \(\mathit{73 }.\mathit{72 }\pm \mathit{6 }.\mathit{47 }\) | \(66.58 \pm 6.28\) | \(70.10 \pm 6.69\) | \(72.95 \pm 6.56\) |
Subject6 | \(\mathit{76 }.\mathit{64 } \pm \mathit{5 }.\mathit{80 }\) | \(66.76 \pm 6.40\) | \(69.18 \pm 6.14\) | \(76.16 \pm 6.30\) |
Subject7 | \(73.92 \pm 6.08\) | \(70.82 \pm 5.53\) | \(70.51 \pm 7.37\) | \(\mathit{74 }.\mathit{76 }\pm \mathit{6 }.\mathit{20 }\) |
PAM | \(73.75 \pm 2.66\) | \(69.12 \pm 2.93\) | \(68.29 \pm 1.19\) | \(\mathit{74 }.\mathit{99 }\pm \mathit{1 }.\mathit{92 }\) |