Fig. 4

Correlations between observed and predicted seed germination (%) based on linear discriminant (LDA) and support vector machine (SVM) classification models. Validation data (see Table 1) were used to predict tomato seed germination (%) in five seed subsamples from two varieties. We performed validations of both linear discriminant (LDA) and support vector machine (SVM) classification models. Seed germination percentages obtained from the seed company are presented as colored circles and considered “known germination”. Blue circles represent germination percentages of samples, which were used as training data. Red colored circles represent germination percentages of validation samples (not included in training data). Colored squares represent predicted germination percentages of training (blue squares) and validation (red squares) samples. Each colored symbol represents germination percentage based on 96 individual seeds