ABSTRACT
Nutrient solution plays an essential role in providing macronutrients to hydroponic plants. Determining nitrogen in the form of nitrate is crucial, as either a deficient or excessive supply of nitrate ions may reduce the plant yield or lead to environmental pollution. This work aims to evaluate the performance of feature reduction techniques and conventional machine learning (ML) algorithms in determining nitrate concentration levels. Two features reduction techniques, linear discriminant analysis (LDA) and principal component analysis (PCA), and seven ML algorithms, for example, k-nearest neighbors (KNN), support vector machine, decision trees, naïve bayes, random forest (RF), gradient boosting, and extreme gradient boosting, were evaluated using a high-dimensional spectroscopic dataset containing measured nitrate-nitrite mixed solution absorbance data. Despite the limited and uneven number of samples per class, this study demonstrated that PCA outperformed LDA on the high-dimensional spectroscopic dataset. The classification accuracy of ML algorithms combined with PCA ranged from 92.7% to 99.8%, whereas the classification accuracy of ML algorithms combined with LDA ranged from 80.7% to 87.6%. The PCA with the RF algorithm exhibited the best performance with 99.8% accuracy.
Subject(s)
Algorithms , Nitrates , Bayes Theorem , Hydroponics , Machine Learning , Support Vector MachineABSTRACT
This research investigates the physicochemical properties of biopolymer succinyl-κ-carrageenan as a potential sensing material for NH4+ Localized Surface Plasmon Resonance (LSPR) sensor. Succinyl-κ-carrageenan was synthesised by reacting κ-carrageenan with succinic anhydride. FESEM analysis shows succinyl-κ-carrageenan has an even and featureless topology compared to its pristine form. Succinyl-κ-carrageenan was composited with silver nanoparticles (AgNP) as LSPR sensing material. AFM analysis shows that AgNP-Succinyl-κ-carrageenan was rougher than AgNP-Succinyl-κ-carrageenan, indicating an increase in density of electronegative atom from oxygen compared to pristine κ-carrageenan. The sensitivity of AgNP-Succinyl-κ-carrageenan LSPR is higher than AgNP-κ-carrageenan LSPR. The reported LOD and LOQ of AgNP-Succinyl-κ-carrageenan LSPR are 0.5964 and 2.7192 ppm, respectively. Thus, AgNP-Succinyl-κ-carrageenan LSPR has a higher performance than AgNP-κ-carrageenan LSPR, broader detection range than the conventional method and high selectivity toward NH4+. Interaction mechanism studies show the adsorption of NH4+ on κ-carrageenan and succinyl-κ-carrageenan were through multilayer and chemisorption process that follows Freundlich and pseudo-second-order kinetic model.
