Classification of basil plant based on the level of consumed nitrogen fertilizer using an olfactory machine

Basil ( Ocimum basilicum ) is an aromatic plant with numerous medical properties that belongs to the Lamiaceae family. The application of nitrogen fertilizer has high impacts on the basil yield improvement. However, using a high amount of nitrogen fertilizer modulates plant metabolisms especially amino acid derivatives and other nitrogen-containing secondary metabolism. Nitrogen fertilizer also increases nitrate accumulation in plant tissues which is unsafe for human health. Therefore, the present study was carried out to classify the basil plant based on the different levels of consumed nitrogen fertilizer using a machine olfaction system (e-nose machine). To achieve this, four different levels of urea fertilizer (0, 50, 100, and 150 kg ha −1 ) were tested. The moisture content of samples was quite similar. Collected data were analyzed by principal component analysis, artificial neural networks, linear discriminant analysis, and quadratic discriminant analysis methods. According to the results, TGS822 sensor displayed the most efficient response for classification. Considering the results of PCA, 90% of data variance was covered by PC1 and PC2. Similarly, confusion matrixes obtained from ANN, LDA, and QDA analysis presented the accuracy of 96.7%, 95%, and 97.78%, respectively. The average values of accuracy, precision, sensitivity, specificity, and AUC parameters per class for the best model of QDA were 0.9891, 0.9638, 0.9638, 0.9937, and 0.9788, respectively.