RandomForestClassifier · 8,000 products · 9 features · 5-class (a–e)
| Model | RandomForestClassifier |
| n_estimators | 100 |
| Train/Test Split | 80/20 |
| Random Seed | 42 |
| Test Accuracy | 0.285 |
| Features | energy, fat, saturated_fat, carbs, sugars, fiber, proteins, salt, sodium |
Note: Low accuracy is expected — the synthetic dataset has a randomly assigned target, independent of features. In a real-world setting, features would have meaningful predictive power for nutrition grade.