Table 5.
Attributes | Regression equation | Performance of calibration | Validation robustness | Sensitivity analysis | |
---|---|---|---|---|---|
R 2 | RMSEC | RMSEV | |||
Firmness |
Y = 1.48 + 1.10*Fp Y = 1.85 + 6.06*Hardness |
0.80 0.72 |
0.9 1.0 |
0.8 0.9 |
2.05 N 0.41 N mm−2 |
Chewiness |
Y = 1.36 + 0.74*Fp Y = 1.39 + 4.52*Hardness |
0.77 0.84 |
0.6 0.5 |
0.6 0.6 |
2.08 N 0.30 N mm−2 |
Stickiness | Y = 6.87 + 59.01*Adhesiveness – 4.92*Springiness | 0.32 | 0.7 | 1.0 | ‐ |
Mealiness | Y = −11.28 + 4.46*Hardness + 17.64*Springiness | 0.82 | 1.0 | 1.2 | 0.56 N mm−2 |
Moist | Y = 7.53–6.48*Hardness | 0.76 | 1.0 | 0.6 | 0.39 N mm−2 |
Calibration samples = 27; Validation samples = 8; Fp: puncture force. The sensitivity analysis provided the minimum instrumental difference required to ensure a perceptible sensory difference (P < 0.1). The table lists the difference only for the first prediction parameter and only for the best predictions.