Near Infrared Spectroscopy Models Research Articles

Understanding the complexity of near-infrared quantification of highly porous patient-tailored drug products by utilizing chemometrics and stimulated Raman imaging.

Additively manufactured drug products, typically produced using small-scale, on-demand batch mode, require rapid and non-destructive quantification methods. A tunable modular design (TMD) approach combining porous polymeric freeze-dried modules and an additive manufacturing method, inkjet printing, was proposed in an earlier study to fabricate accurate and patient-tailored doses of an antidepressant citalopram hydrobromide. This approach addresses the unmet medical needs associated with antidepressant tapering. Non-destructive quantification of printed porous structures is challenging due to the presence of residual solvents and frequent fluctuation of the material density. These shortcomings were mitigated by utilizing a spinning near-infrared spectroscopy (NIRS) measurement setup and a post-print drying step. A machine learning algorithm (ML), specifically support vector regression, was implemented to lessen potential non-linearities caused by the complex structure of TMD drug products. The non-linear support vector regression models performed better than linear partial least squares (PLS) models when modeling the entire sample set (prediction error improved by 19%). By dividing the TMD samples into subtypes and creating individual models for each subtype improved model performance: linear PLS models performed better or equally to non-linear models. It was hypothesized that this outcome was due to the structural differences between different TMD sample subtypes that was later confirmed by stimulated Raman scattering (SRS) microscopy. It was demonstrated that for complex porous drug products ML algorithms can improve NIRS model performance when a single universal robust model is preferred, and SRS is a powerful tool to explain the challenges that printing onto porous drug products can introduce to the NIRS quantification.

A rapid evaluation method for pumpkin sensory quality based on near‐infrared spectroscopy

AbstractThe sensory quality (SQ) of pumpkin fruit is one of the most important quality indexes for consumers. In this study, a rapid pumpkin SQ evaluation method based on near‐infrared spectroscopy (NIRS) was introduced. The starch and moisture contents of pumpkins were first determined by NIRS, and then the SQ of pumpkin could be estimated indirectly by these components. Test was carried out on 97 samples of 34 pumpkin fruits. The pumpkin starch and water NIRS models based on variable dimension particle swarm optimization combined moving window (VDPSO‐CMW) and standard normal variate transformation (SNV) showed the best predictive performances, whose predictive determination coefficients (R2p) were 0.775 and 0.932, respectively. A multiple linear regression model (correlation coefficient = 0.788) was established based on the prediction results of the two models and the SQ values obtained from cooking experiments. The proposed method has the potential to provide effective support for pumpkin breeding and production.

Study on bionics-based swarm intelligence optimization algorithms for wavelength selection in near-infrared spectroscopy

Wavelength selection is one of the most important steps in the modeling of near-infrared spectroscopy (NIRS), which is of great significance to reduce model complexity and improve model performance. In this paper, a total of ten bionics-based swarm intelligence optimization algorithms (BSIOAs) inspired by natural creatures, such as Harris Hawks Optimization (HHO), Butterfly Optimization Algorithm (BOA), Whale Optimization Algorithm (WOA), Monarch Butterfly Optimization (MBO), Grey Wolf Optimization (GWO), Fruit Fly Optimization Algorithm (FOA), Bat Algorithm (BA), Ant Colony Optimization (ACO), Particle Swarm Optimization (PSO), and Genetic Algorithm (GA) were studied on application to wavelength selection in the NIRS modeling. Three benchmark NIRS datasets were used to evaluate the algorithms by calculating the indicators, including coefficients of determination, root mean square error, and residual predictive deviation in calibration and prediction. The results obtained showed that these BSIOAs can significantly reduce the number of wavelengths (retaining half or fewer). Compared with the full-spectrum models, the present models not only simplified the model structures but improved the model performances. The performances were generally better than the ones by some popular and classic wavelength selection algorithms, such as competitive adaptive reweighted sampling, Monte Carlo uninformative variable elimination, variable importance in projection, interval partial least-squares, and successive projections algorithm.

Rapid identification of fragrant rice using starch flavor compound via NIR spectroscopy coupled with GC–MS and Badh2 genotyping

The identification of fragrant rice varieties using near-infrared reflectance spectroscopy (NIRS) models has attracted extensive attention from regulatory authorities worldwide. In this study, 138 fragrant and 54 nonfragrant rice varieties were planted in the same region and distinguished using sensory evaluation, gas chromatography–mass spectrometry analysis, and betaine aldehyde dehydrogenase 2 (Badh2) genotyping. Then, the 2-acetyl−1-pyrroline (2-AP) content was assessed based on partial least-squares discriminant (PLS-DA) models generated after 2nd individually or combined with SNV/MSC/smoothing preprocessing successfully classified fragrant rice both in the calibration and predictive sets. Moreover, design of experiments (DoE)-based preprocessing selection was employed as an effective strategy to optimize the calibration models compared with the one variable at a time (OVAT) method. Further, Badh2 genotype sample screening assisted with identifying authentic fragrant rice and guaranteed the NIRS model's prediction accuracy in identifying fragrant rice. In conclusion, the high throughput PLS-DA multivariate method coupled with NIRS data was applied to identify fragrant rice varieties in routine monitoring and was effective, accurate, and rapid.

Quantitative prediction of minced chicken gel strength under ultrasonic treatment by NIR spectroscopy coupled with nonlinear chemometric tools evaluated using APaRPs

Achieving the ideal gel strength is essential for desired texture in minced chicken products. This study developed a rapid, non-destructive method using near-infrared (NIR) spectroscopy and nonlinear chemometric modeling to predict minced chicken gel strength under ultrasonic treatment. Initially, minced chicken samples were subjected to high-intensity ultrasound for 0–50 min. This was followed by heat-induced gelation. Gel strength was conventionally measured, and NIR spectra were collected. Nonlinearity between gel strength and spectral data was confirmed using augmented partial residual plots (APaRPs). Subsequently, nonlinear support vector machine (SVM) and extreme learning machine (ELM) models were developed using full NIR spectra and variable selection methods, including uninformative variable elimination (UVE), competitive adaptive reweighting (CARS), and genetic algorithms (GA). GA proved most effective for enhancing model performance, achieving the highest predicted coefficient of determination (Rp2 = 0.8772) with the ELM model, demonstrating potential for rapid, non-destructive prediction of minced chicken gel strength quality.

Establishment of near-infrared rapid prediction model and comprehensive evaluation model for foxtail millet quality

Foxtail millet (Setaria italica) is one of the earliest cereal crops in the world and is still an important food crop in Asia, Europe and Africa. However, there are few reports on rapid and comprehensive evaluations of foxtail millet quality. In this study, color, nutrient content, cooking quality and sensory quality of 45 varieties of foxtail millet from the main production areas of the country were determined. Near infrared spectroscopy (NIRS) combined with partial least squares regression (PLSR) was developed for rapid and nondestructive detection of foxtail millet quality, where NIRS models for ash, amylopectin, protein, and VB1 and VB2 were desirable. In addition, correlation analysis, principal component analysis and cluster analysis were used for comprehensive evaluation of millet quality. The results showed that foxtail millet with low ash, protein, fat and fiber content and high amylopectin and carotenoid content has good cooking quality and was more preferred. This research will contribute to the quality evaluation and further processing of foxtail millet.

Physico-chemical and chemometric analysis of milk chocolate sold in Ghana using NIR spectroscopy

Chocolates sold in Ghana are stored under different conditions that are suspected to affect their appearance, flavour and texture. Rapid and non-invasive techniques such as near-infrared spectroscopy (NIRS) have been lauded for their reliability and cost-effectiveness that can be very useful for chocolate monitoring. This study developed a rapid and non-destructive method to predict the quality of chocolate obtained from three different sales outlets based on the location and conditions of retail. Data for physicochemical analysis (total color change, total phenolics, free fatty acid, peroxide value, moisture, hardness, and aluminum content) and mold count of chocolate were collected using standard protocols. These data and results obtained from NIRS in the wavelength range 900–1700 nm, were used to develop chemometric models to predict the parameters measured and classified the chocolate samples. Chocolate from the street recorded the highest mold count of 10.00 ± 18.92 cfu/g. Although the physicochemical analysis showed that different retail conditions had no significant effect on the chocolate quality parameters, the NIRS models could classify the chocolates based on retail conditions, with an average recognition and prediction accuracy of 75.41 % and 71.59 %, respectively. The regression model could predict the total color change with R2CV of 0.503 and RMSECV of 4.96 w/w. The findings suggest that NIRS combined with chemometrics could be used to classify chocolate sold under different conditions at different retail locations. However, the models could not predict other physicochemical quality parameters.

Determination of soluble solids content in loquat using near-infrared spectroscopy coupled with broad learning system and hybrid wavelength selection strategy

Soluble solids content (SSC) serves as a crucial metric in assessing loquat quality. To achieve fast, non-destructive, and accurate detection of SSC, the combination of near-infrared (NIR) spectroscopy and broad learning system (BLS) model was developed in this study. Meanwhile, the influence of six spectral preprocessing methods and six wavelength selection algorithms on SSC prediction was explored, and the performance of the BLS model was compared with the widely used partial least squares regression (PLSR) and least squares support vector machine (LS-SVM) models. The results showed that Savitzky-Golay smoothing combined with standard normal transformation (SG-SNV) offered the optimal pretreatment, and the hybrid method, namely interval variable iterative space shrinking analysis and successive projections algorithm (iVISSA-SPA), emerged as the most effective wavelength selection method. The BLS model outperformed both PLSR and LS-SVM models. Specifically, the BLS model based on the iVISSA-SPA method achieved the optimal SSC prediction results, with RP2 = 0.8646, RMSEP = 0.5104, and RPD = 2.7481. Therefore, NIR spectroscopy coupled with the BLS model and hybrid wavelength selection strategy could rapidly, non-destructively and accurately detect the SSC of loquat, providing a viable alternative for SSC prediction in fruit.

High-throughput near-infrared spectroscopy analysis of nutritional composition in sweet potato stem tips

Leaf-vegetable sweet potato is a nutritious food source, with stem tips as the harvest organ. However, the lack of rapid quantitative approaches for quality evaluation of stem tips largely limits the genetic improvement of leaf-vegetable sweet potato. This study aimed to establish a near-infrared spectroscopy (NIRS) methodology for the rapid analysis of both proximate (cellulose, crude protein, and soluble sugar) and functional (chlorogenic acid, total flavonoid content, and total phenolic content) components in the stem tips. Leveraging a dataset of 140 representative germplasm samples, we developed six robust NIRS models through the optimization of calibration sets and variable selection. These models exhibited exceptional accuracy, with high determination coefficients on calibration (R2C = 0.95–0.98), cross-validation (R2CV = 0.93–0.96), external validation (R2V = 0.91–0.95), and the ratio of prediction to deviation (RPD = 6.78–9.72). Overall, the NIRS models developed through this research facilitate high-throughput profiling of nutritional composition in stem tips, thereby enabling the swift identification of superior germplasm suited for leaf-vegetable sweet potato production.

An efficient method for tracing the geographic origin of Enshi Yulu fresh tea leaves based on near infrared spectroscopy combined with synergy interval PLS and genetic algorithm

This paper applies near-infrared spectroscopy (NIRS) and multiple chemometrics to efficiently distinguish the origins of fresh tea leaves. The key components were obtained using the partial least squares discriminant analysis (PLS-DA) method. PLS, synergy interval PLS (siPLS), principal component analysis (PCA), genetic algorithm (GA), and their combination methods were used to establish NIRS non-destructive discrimination models. Then, the practical application was examined using external samples. The study identified nine key components (variable importance for the projection (VIP) > 1): epigallocatechin, epicatechin, total sugar, water extracts, total catechins, gallocatechin gallate, tea polyphenols, gallocatechin, and epigallocatechin gallate. Of the six NIRS models, the siPLS-GA model that used 37 spectral data points produced the best results (Rp2 = 0.9706, RMSEP = 0.0772, RPD = 6.59). This model had a prediction accuracy of 96.67% for the prediction set samples and 93.33% for the external samples. It offers a rapid, precise, and non-invasive approach to monitor and regulate the illicit trade of fresh tea leaves, thereby guaranteeing the authenticity of Enshi Yulu products from the processing source and fostering the long-term prosperity and stability of the Enshi Yulu tea industry.

Contrasted effects of shade tree legumes on soil organic carbon stock and carbon balance in 20-year cacao agroforestry, Ivory Coast

Cacao–based agroforestry systems are promoted as adaptation and mitigation solutions for cacao production and carbon sequestration. Based on a 20-year experiment, we assessed the impact of associated shade tree legume (ATL), Albizia lebbeck and Acacia mangium on the total carbon stock (in soil at 60 cm depth + tree biomass + litter) of cacao stands. This study included cacao systems shaded with either A. lebbeck (Cacao-Alb) or A. mangium (Cacao-Aca) and full-sun cacao stands (Control). Soil organic carbon (SOC) contents (up to 60 cm deep) were estimated by a calibrated near-infrared spectroscopy model. Total tree biomasses were estimated using allometric equations. Leaf litter was sampled from 1-m2 quadrats. Compared to Control, Cacao-Aca had a significant negative impact on the carbon stock in the cacao biomass (−47%) as well and in the soil at depths of 10 cm (−23%), 30 cm (−21%) and 60 cm (−12%). In contrast, Cacao-Alb had a nonsignificant effect on carbon storage in the cacao biomass, whereas it generally had a positive influence on the SOC stock regardless of depth, i.e., +6% at the 0–10 cm depth, +7% at 0–30 cm, +20% at 30–60 cm and + 11% at 0–60 cm. Cacao-Aca had a significant positive impact (+71%) on the total carbon stock per hectare. The increase in Cacao-Alb relative to that in the Control reached +38%, but the difference was not significant. These contrasting results between the two tree legume species could be explained by the high-quality litter, reflected by the lower C/N and C/P ratios produced by A. lebbeck, and the greater negative impact of A. mangium on cacao biomass. The main finding of this study is that the impact of intercropping cacao with shade tree legumes on the stand-level total carbon stock depends on the ATL species.

Assessing the Application of Near-Infrared Spectroscopy to Determine Saccharification Efficiency of Corn Biomass

Nowadays, in the bioethanol production process, improving the simplicity and yield of cell wall saccharification procedure represent the main technical hurdles to overcome. This work evaluated the application of a rapid and cost-effective technology such as near -infrared spectroscopy (NIRS) for easily predict saccharification efficiency from corn stover biomass. Calibration process focussing on the number of samples and the genetic background of the maize inbred lines were tested; while Modified Partial Least Squares Regression (MPLS) and Multiple Linear Regression (MLR) were assessed in predictions. The predictive capacity of the NIRS models was mainly determined by the coefficient of determination (r2ev) and the index of prediction to deviation (RPDev) in external validation. Overall, we could check a better efficiency of the NIRS calibration process for saccharification using larger number of observations (1500 sample set) and genetic backgrounds; while MPLS regression provided better prediction statistics (r2ev = 0.80; RPDev = 2.21) compared to MLR (r2ev = 0.68; RPDev = 1.75). These results indicate that NIRS could be successfully implemented as a large-phenotyping tool in order to test the saccharification potential of corn biomass.

Model Exploration and Application of Near-Infrared Spectroscopy for Species Separation and Quantification during Mixed Litter Decomposition in Subtropical Forests of China

Litter of different species coexists in the natural ecosystem and may induce non-additive effects during decomposition. Identifying and quantifying the origins of species in litter mixtures is essential for evaluating the responses of each component species when mixed with co-occurring species and then unraveling the underlying mechanism of the mixing effects of litter decomposition. Here, we used near-infrared spectroscopy (NIRS) to predict the species composition and proportions of four-tree species foliage mixtures in association with litter crude ash and litter decomposition time. To simulate the whole mixed litter decomposition process in situ, a controlled mixture of four tree species litter leaves consisting of 15 tree species combinations and 193 artificial mixed-species samples were created for model development and verification using undecomposed pure tree species and decomposed litter of single tree species over one year. Two series of NIRS models were developed with the original mass and ash-free weight as reference values. The results showed that these NIRS models could provide an accurate prediction for the percentage of the component species from in the litter leaf mixture’s composition. The predictive ability of the near-infrared spectroscopy model declined marginally with the prolonged litter decomposition time. Furthermore, the model with ash-free litter mass as a reference exhibited a higher coefficient of determination (R2) and a lower standard error of prediction (RMSECV). Thus, our results demonstrate that NIRS presents great potential for not only predicting the organic composition and proportion in multi-species mixed samples in static conditions, but also for samples in dynamic conditions (i.e., during the litter decomposition process), which could facilitate evaluation of the species-specific responses and impacts on the interspecific interactions of co-occurring species in high-biodiversity communities.

Near-Infrared Spectroscopy Modeling of Combustion Characteristics in Chip and Ground Biomass from Fast-Growing Trees and Agricultural Residue

This study focuses on the investigation and comparison of combustion characteristic parameters and combustion performance indices between fast-growing trees and agricultural residues as biomass sources. The investigation is conducted through direct combustion in an air environment using a thermogravimetric analyzer (TGA). Additionally, partial least squares regression (PLSR)-based models were developed to assess combustion performance indices via near-infrared spectroscopy (NIRS), serving as a non-destructive alternative method. The results obtained through the TGA reveal that, specifically, fast-growing trees display higher average ignition temperature (227 °C) and burnout temperature (521 °C) in comparison to agricultural residues, which exhibit the values of 218 °C and 515 °C, respectively. Therefore, fast-growing trees are comparatively difficult to ignite, but sustain combustion over extended periods, yielding higher temperatures. However, despite fast-growing trees having a high ignition index (Di) and burnout index (Df), the comprehensive combustion performance (Si) and flammability index (Ci) of agricultural residue are higher, indicating the latter possess enhanced thermal and combustion reactivity, coupled with improved combustion stability. Five distinct PLSR-based models were developed using 115 biomass samples for both chip and ground forms, spanning the wavenumber range of 3595–12,489 cm−1. The optimal model was selected by evaluating the coefficients of determination in the prediction set (R2P), root mean square error of prediction (RMSEP), and RPD values. The results suggest that the proposed model for Df, obtained through GA-PLSR using the first derivative (D1), and Si, achieved through full-PLSR with MSC, both in ground biomass, is usable for most applications, including research. The model yielded, respectively, an R2P, RMSEP, and RPD, which are 0.8426, 0.4968 wt.% min⁻4, and 2.5; and 0.8808, 0.1566 wt.%2 min⁻2 °C⁻3, and 3.1. The remaining models (Di in chip and ground, Df, and Si in chip, and Ci in chip and ground biomass) are primarily applicable only for rough screening purposes. However, including more representative samples and exploring a more suitable machine learning algorithm are essential for updating the model to achieve a better nondestructive assessment of biomass combustion behavior.

Non-Destructive Prediction of Moisture and Fat Content in Cocoa Beans Using Near-Infrared Spectroscopy and Multivariate Regression Models

Non-Destructive Prediction of Moisture and Fat Content in Cocoa Beans Using Near-Infrared Spectroscopy and Multivariate Regression Models

Fast and non- destructive multivariate test method to predict bread wheat grain major quality parameters

ABSTRACT For decades researches have been conducted in Ethiopia to improve wheat productivity and quality. This involved extensive selection of breeding lines based on agronomic and laboratory data from numerous trials. However, generating laboratory data for all these services was both time-consuming and expensive. Hence, the study aimed to develop a very rapid and cost-effective testing method for wheat genotypes and varieties using near-infrared spectroscopy. For this purpose, we collected 155 bread wheat samples from high-growing potential areas of Sinana, Kulumsa, and Holeta. The wet chemistry analysis was done in duplicate following the international standard official methods. The same samples were also scanned using TANGO’S Bruker NIRS machine in duplicate to acquire spectral data. A calibration model was then developed by fitting the reference and spectral data using the partial least square (PLS). The calibration models exhibited a high level of strength for all parameters. The performance of the calibration model for protein (R2 = 0.99; RPD = 9.63), moisture (R2 = 0.96; RPD = 5.29), ash (R2 = 0.98; RPD = 6.39), total gluten (R2 = 0.98; RPD = 8.09), dry gluten (R2 = 0.97; RPD = 5.57), gluten index (R2 = 0.96; RPD = 5.23), and falling number (R2 = 0.97 and RPD = 6.23) were deemed suitable for wheat process and quality control purposes. But the models for moisture and gluten index (R2 = 0.96) were found to be suitable for screening and simple quality control purposes only. Therefore, this study clearly demonstrated that the near-infrared spectroscopy model is a highly reliable and robust method for predicting the key quality traits of bread wheat grains, serving the needs of both breeding programs and industry raw material control.

A new application of Elasticnet regression based near-infrared spectroscopy model: Prediction and analysis of 2,3,5,4′-tetrahydroxy stilbene-2-O-β-D-glucoside and moisture in Polygonum multiflorum

This study constructed a near-infrared spectroscopy (NIRS) quantitative analysis model of 2,3,5,4′-tetrahydroxy stilbene-2-O-β-D-glucoside (TSG) and Polygonum multiflorum's water content based on Elasticnet regression (ER). NIRS data, TSG and the moisture content data sets of 117 samples of P. multiflorum were obtained by using NIR scanning and high-performance liquid chromatography analysis (HPLC) methods. MSE was used to determine the hyperparameters α and λ of Elasticnet. The optimal α values of TSG and water content were 0.83 and 0.01, and the optimal λ values were 0.03 and 1.13, respectively. By preprocessing the original spectral data, the NIRS model was established based on the source data set using ER, compared with the model established by PLS and its optimization algorithm. The results show that the ER model has good performance compared with other methods, with R2c value was 0.9879, RMSEC value was 0.09, R2p value was 0.9784, RMSEP value was 0.13 and RPD value was 6.8. In addition, according to the properties of Elasticnet's variable selection, we speculate that the final coefficients of spectral features may indirectly reflect the spectral features of different components. The results showed better prediction accuracy and superior performance. This is a much more reliable method used to forecast the content of effective components in plants.

Development and validation of a near-infrared spectroscopy model for the prediction of muscle protein in Chinese native chickens

This study investigated the ability of the near-infrared spectroscopy (NIRS) model to predict the protein of freeze-dried muscle samples in Chinese native chickens and to determine the accuracy of the models for other native chicken breeds. Spectral pretreatment, wavelength selection, and outlier sample elimination were used to optimize the calibration models. The results showed that the best model was obtained by using a combination of standard normal variable transformation and gap-segment first-derivative pretreatment spectra after removing 48 outliers in the wavelength range of 1,439 to 1,900 nm, with coefficient of determination for the calibration (R2C) of 0.95, standard error of cross-validation (SECV) of 1.18, coefficient of determination for the prediction (R2P) of 0.95, the ratio of the standard deviation of the validation to the standard deviation of the calibration (RPDP) of 4.62. The findings indicated that NIRS can be used to predict the protein of freeze-dried muscle in Chinese native chickens.

Exploring near-infrared spectroscopy and hyperspectral imaging as novel characterization methods for anaerobic gut fungi.

Neocallimastigomycota are a phylum of anaerobic gut fungi (AGF) that inhabit the gastrointestinal tract of herbivores and play a pivotal role in plant matter degradation. Their identification and characterization with marker gene regions has long been hampered due to the high inter- and intraspecies length variability in the commonly used fungal marker gene region internal transcribed spacer (ITS). While recent research has improved methodology (i.e. switch to LSU D2 as marker region), molecular methods will always introduce bias through nucleic acid extraction or PCR amplification. Here, near-infrared spectroscopy (NIRS) and hyperspectral imaging (HSI) are introduced as two nucleic acid sequence-independent tools for the characterization and identification of AGF strains. We present a proof-of-concept for both, achieving an independent prediction accuracy of above 95% for models based on discriminant analysis trained with samples of three different genera. We further demonstrated the robustness of the NIRS model by testing it on cultures of different growth times. Overall, NIRS provides a simple, reliable, and nondestructive approach for AGF classification, independent of molecular approaches. The HSI method provides further advantages by requiring less biomass and adding spatial information, a valuable feature if this method is extended to mixed cultures or environmental samples in the future.

A Pre-Trial Study to Identify Species of Origin in Halloumi Cheese Utilising Chemometrics with Near-Infrared and Hyperspectral Imaging Technologies

Halloumi cheese has recently gained a Protected Designation of Origin (PDO) indicator, which is related to the place (Cyprus) in which halloumi cheese is made. The PDO label is linked with several requirements, e.g., milk species, quantities, etc.; thus, it is important to study this product regarding authenticity. The utility of using two spectroscopic techniques, hyperspectral imaging (HSI) (400–1000 nm) and conventional near-infrared spectroscopy (NIR) (800–2500 nm) were assessed for the discrimination of 17 Cypriot halloumi cheese types, which could be categorized as of cow or goat–sheep origin. The aim of this study was to obtain spectral information for halloumi cheese using other promising infrared and imaging spectroscopic techniques as a comparison to a previously acquired mid-infrared (MIR) spectroscopy dataset. NIR and HSI are both fast and easy techniques in application, both of which provide significant information in food analysis. Chemometric analysis was crucial for interpreting the spectroscopic data by applying the unsupervised methods: principal component analysis (PCA) and hierarchical cluster analysis (HCA). The HSI model was found to be based intuitively on the appearance of cheese samples after freeze-drying (e.g., color; yellow/white, and texture; oily/dry), while the NIR grouping of samples was determined to be based on composition, mainly fat, protein and lactose content of the cheese samples. The HSI model returned distinct clusters of the two halloumi cheese types, cow and goat–sheep origin, with one outlier (16/17 accuracy; 94%), while the NIR model proved less accurate (13/17; 76%).