Table 3. Predicted end-use class-membership probabilities for four additional commercial strong wheat flours
| Probability for lean fermented dough (P1) | Probability for non-fermented dough (P2) | Probability for high-formula fermented dough (P3) |
| Sample A | 0.5966 | 0.3789 | 0.0244 |
| Sample B | 0.5423 | 0.3971 | 0.0606 |
| Sample C | 0.6413 | 0.2845 | 0.0742 |
| Sample D | 0.3508 | 0.5776 | 0.0716 |
The values represent the predicted probabilities calculated by the multinomial logistic regression model. For each sample, the sum of probabilities (P1+P2+P3) equals 1.0. The class with the highest probability was presented as the model- predicted primary end-use candidate. These probabilities represent relative class membership within the scope of the training data and should not be interpreted as absolute probabilities of product success (n=4).
P1: Probability for lean fermented dough; P2: Probability for non-fermented dough; P3: Probability for high-formula fermented dough.