Machine learning-based prediction of total phenolic and flavonoid in horticultural products

Kusumiyati Kusumiyati,Yonathan Asikin

DOI: https://doi.org/10.1515/opag-2022-0163

2023-01-01

Open Agriculture

Abstract:Abstract The purpose of this study was to predict the total phenolic content (TPC) and total flavonoid content (TFC) in several horticultural commodities using near-infrared spectroscopy (NIRS) combined with machine learning. Although models are typically developed for a single product, expanding the coverage of the model can improve efficiency. In this study, 700 samples were used, including varieties of shallot, cayenne pepper, and red chili. The results showed that the TPC model developed yielded R 2 cal, root mean squares error in the calibration set, R 2 pred, root mean squares error in prediction set, and ratio of performance to deviation values of 0.79, 123.33, 0.78, 124.20, and 2.13, respectively. Meanwhile, the TFC model produced values of 0.71, 44.52, 0.72, 42.10, and 1.87, respectively. The wavelengths 912, 939, and 942 nm are closely related to phenolic compounds and flavonoids. The accuracy of the model in this study produced satisfactory results. Therefore, the application of NIRS and machine learning to horticultural products has a high potential of replacing conventional laboratory analysis TPC and TFC.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to use near - infrared spectroscopy (NIRS) combined with machine - learning techniques to predict the total phenolic content (TPC) and total flavonoid content (TFC) in several horticultural products. Traditionally, the determination of these quality attributes requires chemical analysis in the laboratory, which is not only time - consuming and costly, but also generates chemical waste that is harmful to the environment. Therefore, researchers hope to develop a rapid and non - destructive method to replace the traditional laboratory analysis method in order to improve efficiency and reduce environmental pollution. Specifically, the researchers used 700 samples, including different varieties of onions, peppers and red peppers, collected data through NIRS, and combined multiple machine - learning algorithms to establish prediction models. The goal of the study was to verify whether the combination of NIRS and machine - learning can effectively predict the TPC and TFC of these horticultural products, and which spectral pre - processing method can best improve the prediction ability of the model. The research results show that the original spectral data performs best in predicting TPC and TFC. Although multiple spectral pre - processing methods were tried, they did not significantly improve the accuracy of the model. This may be because the data set itself has high variability, and spectral pre - processing usually functions by minimizing the variability of data. In addition, the study also found that the 900 - 950 - nanometer wavelength region is particularly important for establishing prediction models for TPC and TFC.

Machine learning-based prediction of total phenolic and flavonoid in horticultural products

Nondestructive prediction of total phenolics, flavonoid contents, and antioxidant capacity of rice grain using near-infrared spectroscopy

Rapid Quantitative Determination of Multi-Indicators in Lonicerae Japonicae Flos by Near-Infrared Spectroscopy

Non-destructive prediction of total soluble solids in strawberry using near infrared spectroscopy

Synchronously Predicting Tea Polyphenol and Epigallocatechin Gallate in Tea Leaves Using Fourier Transform-Near-Infrared Spectroscopy and Machine Learning

Classification and Prediction by Pigment Content in Lettuce (Lactuca sativa L.) Varieties Using Machine Learning and ATR-FTIR Spectroscopy

Determination of Proanthocyanidins and Total Phenolics in the Seedpod of Nelumbo nucifera Gaertn by Near Infrared Diffuse Reflectance Spectroscopy

Application of Near Infrared Spectroscopy Combined with Particle Swarm Optimization Based Least Square Support Vactor Machine to Rapid Quantitative Analysis of Corni Fructus

Machine learning approach for high-throughput phenolic antioxidant screening in black Rice germplasm collection based on surface FTIR

Non-destructive measurement of antioxidant activity and water content in chili powder (Capsicum annuum L.) using near-infrared spectroscopy

Development of Predictive Models for Total Phenolics and Free P-Coumaric Acid Contents in Barley Grain by Near-Infrared Spectroscopy.

Phenolic content discrimination in Thai holy basil using hyperspectral data analysis and machine learning techniques

Measurement of Chlorophyll Content in Pepper Leaves by Near Infrared Analysis

An Approach to Predicting Antioxidative Activities of Natural Products Based on Near Infrared Spectroscopy

Nondestructive Prediction of Isoflavones and Oligosaccharides in Intact Soybean Seed Using Fourier Transform Near-Infrared (FT-NIR) and Fourier Transform Infrared (FT-IR) Spectroscopic Techniques

VIS-NIR Modeling of Hydrangenol and Phyllodulcin Contents in Tea-Hortensia (Hydrangea macrophylla subsp. serrata)

Developing Prediction Models Using Near-Infrared Spectroscopy to Quantify Cannabinoid Content in Cannabis Sativa

Rapid In Situ Near-Infrared Assessment of Tetrahydrocannabinolic Acid in Cannabis Inflorescences before Harvest Using Machine Learning

Machine learning methods in near infrared spectroscopy for predicting sensory traits in sweetpotatoes

Research on New Methods to Obtain Plant Growth Information in Facility Agriculture by Near-Infrared Spectrum Analysis

Determination of antioxidant capacity and phenolic content of haskap berries (Lonicera caerulea L.) by attenuated total reflectance-Fourier transformed-infrared spectroscopy