Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Development of a Near Infrared Spectroscopy Model for Prediction of Fibre Compounds in Alfalfa

Andersen, Christina LU (2020) KMBM05 20201
Applied Microbiology
Abstract
Background: This project investigates if it is possible to develop a calibration model from near infrared (NIR) spectroscopic measurements, for determination of the amount and type of fibre fractions within protein powder produced from the legume alfalfa, without performing wet experiments. Alfalfa is also known as Medicago sativa and lucerne, but is in this project further referred to as alfalfa. Such a model would be applicable as a protein powder production process control, by scanning a small amount of sample during the production process, immediately resulting in a fibre content value. With this result, one will know when the process should be stopped by means of nutritional values. Except from fibres, alfalfa contains large amounts... (More)
Background: This project investigates if it is possible to develop a calibration model from near infrared (NIR) spectroscopic measurements, for determination of the amount and type of fibre fractions within protein powder produced from the legume alfalfa, without performing wet experiments. Alfalfa is also known as Medicago sativa and lucerne, but is in this project further referred to as alfalfa. Such a model would be applicable as a protein powder production process control, by scanning a small amount of sample during the production process, immediately resulting in a fibre content value. With this result, one will know when the process should be stopped by means of nutritional values. Except from fibres, alfalfa contains large amounts of nutrients, for example essential amino acids. The advantageous amino acids are
thus extracted from the fibrous alfalfa during the protein powder production process. The alfalfa protein powder is produced from stems, leaves and flowers of intact, freshly harvested alfalfa plants. The raw alfalfa was frozen during storage, then thawed and wetted prior to the first press, which is resulting in a protein rich green juice, and a fibrous pulp. The pH of the green juice is decreased to precipitate proteins. The green juice is then centrifuged resulting in a pellet consisting of the total water soluble solid content extracted from alfalfa. The pellet is freeze dried into a protein powder in order to concentrate the protein content. This process is performed 10 times in
total, the first time untreated raw wetted alfalfa is pressed into green juice as mentioned, the following nine times the fibrous pulp from the prior press is wetted and pressed into new samples of green juice. The aim of re-pressing the fibrous pulp is to extract the highest total amount of protein from one batch of alfalfa. This protein powder production from raw untreated alfalfa to protein powder, does increase human digestibility of alfalfa
by increasing the amount of protein per weight. Protein powder derived from each of the 10 presses was collected in separate fractions to determine to which extent the fibre profile is changing using an enzymatic gravimetric method. The amounts of protein, insoluble dietary fibre (IDF), soluble dietary fibre (SDF), total dietary fibre (TDF), available carbohydrates (ACH) and ash were determined, since NIR spectra are affected by all compounds of the protein powder. NIR spectra from all 10 presses are related directly to the determined TDF contents, which are used as reference values in order to calibrate a partial least squares (PLS) model that produces predicted TDF values. Attempts were also made to conduct NIR spectra earlier in the protein powder production process, from the green juice prior to centrifugation and from the pellet prior to freeze drying. A cellulose gluten powder dilution series comparable to
the 10 presses of protein powder was prepared, to test if a calibration model could be developed from NIR spectral data of powder containing cellulose as one of the main components. The cellulose gluten spectra were also compared with
protein powder spectra during spectral compound analyses.

Results: The nutrient profile determination resulted in a total decreasing amount of protein from 43.12% w/w for press 1 to 37.84% w/w for press 10. The TDF content increased from 22.80% w/w for press 1 to 47.47% w/w for press 10. ACH decreased from 5.43% w/w for press 1 to 1.10% w/w for press 10, while the amount of determined ash decreased from 8.24% w/w for press 1 to 2.70% w/w for press 10. Usable and promising NIR spectra were conducted from all measured protein and cellulose gluten powder samples. A calibration model predicting TDF contents for each of the 10 presses was developed with a wavenumber range from 6,800 cm-1 to 4,100 cm-1 and R2 = 0.98. For all 10 presses, the mean deviation from the reference TDF contents was 0.76% w/w. NIR spectra from the green juice and pellet could not be conducted with the available NIR instrument and presetting options.

Conclusions: It is challenging to convert complex NIR spectra into usable information. Since a broad wavenumber spectrum was chosen for the model development, it was easy to fit the spectra to almost any kind of reference
values, even though the spectra do not describe those reference values. It also has to be kept in mind that the model is not validated. Therefore it is hard to draw conclusions regarding the model quality. It can be concluded though, that
NIR spectra obtained from the protein powder of alfalfa look promising for further investigation, since a good correlation between the TDF amounts and NIR spectra could be seen. Of future work the first priority should be to validate this produced model. If that looks promising, both a new independent validation set and a larger data set to produce a new calibration model is required to further test the model robustness. (Less)
Abstract (Swedish)
Bakgrund: Detta projekt undersöker om det är möjligt att utveckla en kalibreringsmodell utifrån spektroskopiska mätningar i det nära infraröda (NIR) området, för bestämning av mängden och typen av fiberfraktioner i proteinpulver producerat från baljväxten alfalfa, utan att utföra våtexperiment. Alfalfa benämnas även Medicago sativa och lucerne, men kallas inom detta projekt alfalfa. En sådan typ av modell skulle kunna tillämpas som en kontroll av proteinpulvrets produktionsprocess, genom att skanna en liten mängd prov från ett steg i produktionsprocessen, vilket omedelbart resulterar i ett fiberinnehållsvärde. Med ett sådant resultat får man reda på när processen ska stoppas enligt näringsvärdena. Förutom fibrer, innehåller alfalfa en hög... (More)
Bakgrund: Detta projekt undersöker om det är möjligt att utveckla en kalibreringsmodell utifrån spektroskopiska mätningar i det nära infraröda (NIR) området, för bestämning av mängden och typen av fiberfraktioner i proteinpulver producerat från baljväxten alfalfa, utan att utföra våtexperiment. Alfalfa benämnas även Medicago sativa och lucerne, men kallas inom detta projekt alfalfa. En sådan typ av modell skulle kunna tillämpas som en kontroll av proteinpulvrets produktionsprocess, genom att skanna en liten mängd prov från ett steg i produktionsprocessen, vilket omedelbart resulterar i ett fiberinnehållsvärde. Med ett sådant resultat får man reda på när processen ska stoppas enligt näringsvärdena. Förutom fibrer, innehåller alfalfa en hög andel essentiella aminosyror. Aminosyrorna skall därför gärna extraheras från den fibrösa alfalfa under produktionen av proteinpulver.
Proteinpulvret produceras från stjälkar, blad och blommor av intakta, nyskördade alfalfa baljväxter. Den råa alfalfa har av lagringsskäl frysts ner. Frusen alfalfa har därför tinats, fuktats och blivit pressad med en skruvpress, vilket resulterade i en proteinrik grön juice och en fibrös massa. pH-värdet i den gröna juicen sänktes för att fälla ut proteiner. Den gröna juicen centrifugeras sedan, vilket resulterade i en pellets bestående av det totala vattenlösliga fasta innehållet extraherat från alfalfa. Pelleten frystorkades till ett proteinpulver för att koncentrera upp proteininnehållet. Denna process utfördes totalt 10 gånger, första gången med rå alfalfa som pressades till grön juice, följande nio gånger återfuktades den fibrösa massan från den tidigare pressen och pressades sedan till nya separata prover av grön juice.
Syftet med att pressa den fibrösa massan från den tidigare pressen, är att extrahera den högsta totala mängden protein från en batch av alfalfa. Detta sätt att producera proteinpulver på, ökar smältbarheten av alfalfa genom att öka mängden protein per vikt. Proteinpulver från var och en av de 10 pressarna uppsamlas i separata fraktioner för att bestämma i vilken utsträckning fiberprofilen förändrades med användning av en enzym metod. Mängderna protein, olösliga kostfibrer (IDF), lösliga kostfibrer (SDF), totala kostfibrer (TDF), tillgängliga kolhydrater (ACH) och ask bestämdes, eftersom NIR-spektra påverkas av alla föreningar i proteinpulvret. NIR-spektra från alla 10 pressar relaterades direkt till det bestämda TDF-innehållet, som används som referensvärden för att kalibrera en partial least squares (PLS) modell, som i sin tur skall producera förutsagda TDF-värden.
Försök gjordes också för att möjliggöra mätning av NIR-spektra tidigare i proteinpulvrets produktionsprocess. NIR-spektra från den gröna juicen innan centrifugering, och från pelleten före frystorkning försöktes mätas. En utspädningsserie av cellulosa-glutenpulver jämförbar med de 10 pressarna av proteinpulver framställdes för att testa om en kalibreringsmodell kunde utvecklas utifrån NIR-spektra med pulver innehållande cellulosa som en av huvudkomponenterna. Cellulosa-glutenspektra jämfördes också med proteinpulver spektra för att jämföra förekomsten av kemiska föreningar i spektrumen.

Resultat: Bestämning av näringsprofilen resulterade i en total minskande mängd protein från 43,12% w/w för press 1 till 37,84% w/w för press 10. TDF-innehållet ökade från 22,80% w/w för press 1 till 47,47% w/w för press 10. ACH minskade från 5,43% w/w för press 1 till 1,10% w/w för press 10, medan mängden bestämd ask minskade från 8,24% w/w för press 1 till 2,70% w/w för press 10. Användbara och lovande NIR-spektra togs fram för alla uppmätta prover för både proteinpulver och cellulosa-glutenpulver. En kalibreringsmodell som producera förutsagda TDF-värden för var och en av de 10 pressarna utvecklades från vågtalet 6.800 cm-1 till 4.100 cm-1 med R2 = 0,98. För alla 10 pressar var den genomsnittliga medelavvikelsen från referens-TDF-innehållet 0,76% w/w. Uppmätta NIR-spektra med det tillgängliga NIR-instrument och dess förinställningsalternativ från grön juice och pellet kunde inte användas.

Slutsatser: Det är svårt att konvertera komplexa NIR-spektra till användbar information. Eftersom ett brett vågtalsspektrum valdes för modellutvecklingen, är det enkelt att anpassa spektra till nästan alla typer av referensvärden, även om spektrumen inte beskriver just dessa referensvärden. Modellen är inte validerad, och det är därför svårt att dra slutsatser angående modellkvaliteten. Slutsatsen att NIR-spektra erhållna från proteinpulvret i alfalfa ser lovande ut för ytterligare undersökningar, kan dock dras, eftersom en god korrelation mellan TDF-värdena och NIR-spektrumen kunde ses. För framtida arbete bör första prioritet vara att validera denna producerade modell. Om det ser lovande ut, krävs både en oberoende validering av denna modell, och sedan en större datauppsättning för att producera en ny kalibreringsmodell, för att öka modellens robusthet. (Less)
Abstract (Danish)
Baggrund: Dette projekt undersøger om det er muligt at udvikle en kalibreringsmodel ud fra spektroskopiske målinger i det nær infrarøde (NIR) område, for at bestemme mængden og typen af fiberfraktioner i proteinpulver produceret fra bælgplanten alfalfa, uden at udføre klassiske kemiske laboratorieeksperimenter. Alfalfa, også kendt under navnene Medicago sativa og lucerne, vil fortsat refereres til som alfalfa. En sådan model kan anvendes som proceskontrol ved at scanne en lille mængde proteinpulverprøve under produktionsprocessen, hvilket umiddelbart vil resultere i en fiberindholdsværdi. Dette resultat vil give en indikation af, hvornår processen skal stoppes i forhold til proteinpulverets næringsværdier. Udover fibre indeholder alfalfa... (More)
Baggrund: Dette projekt undersøger om det er muligt at udvikle en kalibreringsmodel ud fra spektroskopiske målinger i det nær infrarøde (NIR) område, for at bestemme mængden og typen af fiberfraktioner i proteinpulver produceret fra bælgplanten alfalfa, uden at udføre klassiske kemiske laboratorieeksperimenter. Alfalfa, også kendt under navnene Medicago sativa og lucerne, vil fortsat refereres til som alfalfa. En sådan model kan anvendes som proceskontrol ved at scanne en lille mængde proteinpulverprøve under produktionsprocessen, hvilket umiddelbart vil resultere i en fiberindholdsværdi. Dette resultat vil give en indikation af, hvornår processen skal stoppes i forhold til proteinpulverets næringsværdier. Udover fibre indeholder alfalfa en høj andel essentielle aminosyrer. Disse essentielle aminosyrer ekstraheres fra den fiberholdige alfalfa under produktionsprocessen af proteinpulver.
Proteinpulveret er produceret af stængler, blade og blomster fra intakte, friskhøstede alfalfa-planter. Friskhøstet alfalfa blev nedfrosset for at muliggøre en længere opbevaringstid. Ved projektets start, blev den frosne alfalfa tøet op, fugtiggjort og presset til en proteinrig grøn juice med en skruepresse, hvilket desuden resulterede i en fiberholdig grøn pulp. pH i den grønne juice sænkes for at udfælde proteinerne. Derefter centrifugeres den grønne juice, hvilket resulterer i en pellet bestående af det samlede vandopløselige faste indhold ekstraheret fra alfalfa. Pelleten frysetørres til et proteinpulver for at opkoncentrere de ønskede proteiner. Denne proces udføres i alt 10 gange. Første gang udføres den med optøet ubehandlet alfalfa, og de følgende ni gange med den fiberholdige grønne pulp fra det forudgående pres, der fugtiggøres og presses til nye prøver af grøn juice.
Målet ved at genpresse den fiberholdige grønne pulp, er at ekstrahere den højeste samlede mængde protein fra en portion presset alfalfa. Denne slags proteinpulverproduktion øger fordøjeligheden af alfalfa ved at øge mængden af ​​protein pr. vægt. Proteinpulver afledt fra hver af de 10 pres blev opsamlet i separate fraktioner for at bestemme i hvilket omfang fiberprofilen ændrede sig ved anvendelse af en enzymatisk gravimetrisk metode. Mængderne af protein, uopløselige kostfibre (IDF), opløselige kostfibre (SDF), totale kostfibre (TDF), tilgængelige kulhydrater (ACH) og aske blev undersøgt, da NIR-spektre påvirkes af hele proteinpulverets indhold. NIR-spektre fra alle 10 pres blev relateret direkte til det undersøgte TDF-indhold, der bruges som referenceværdier for at kalibrere en partial least squares (PLS) model, der forudsiger TDF-værdier. \\
Det blev også forsøgt at muliggøre måling af NIR-spektre tidligere i proteinpulverproduktionsprocessen. Der blev forsøgt at måle NIR-spektre af den grønne juice før centrifugering, og på pelleten før frysetørring. En fortyndingsserie med celluloseglutenpulver sammenlignelig med proteinpulver fra de 10 pres af alfalfa, blev fremstillet for at teste, om der kunne udvikles en kalibreringsmodel ud fra NIR-spektre målt på pulver med cellulose som en af hovedbestandsdelene. Celluloseglutenspektrene blev også sammenlignet med proteinpulverspektrene for at sammenligne tilstedeværelsen af ​​kemiske forbindelser i spektrene.

Resultater: Den enzymatiske næringsprofilbestemmelse resulterede i en total faldende mængde protein fra 43,12% w/w for pres 1 til 37,84% w/w for pres 10. TDF-indholdet steg fra 22,80% w/w for pres 1 til 47,47% w/w for pres 10. ACH faldt fra 5,43% w/w for pres 1 til 1,10% w/w for pres 10, mens mængden af aske faldt fra 8,24% w/w for pres 1 til 2,70% w/w for pres 10. Brugbare og lovende NIR-spektre fra alle målte protein- og cellulosegluten-pulverprøver blev indsamlet. En kalibreringsmodel, der forudsagde TDF-indhold for hvert af de 10 pres, blev udviklet med et bølgetalsområde fra 6.800 cm-1 till 4.100 cm-1 med R2 = 0,98. For alle 10 pres var den gennemsnitlige middelafvigelse fra reference-TDF-indholdet 0,76% w/w. NIR-spektre med det tilgængelige NIR-instrument og forindstillingsmulighederne fra den grønne juice og pellet var ikke brugbare.

Konklusioner: Det er svært at konvertere komplekse NIR-spektre til brugbar information. Eftersom der blev valgt et bredt bølgetalspektrum til modeludviklingen, er det potentielt set let at tilpasse spektrene til enhver form for referenceværdier, selvom spektrene ikke beskriver disse referenceværdier. Da modellen ikke er valideret er det svært at drage endelige konklusioner vedrørende modelkvaliteten. Det kan imidlertid konkluderes, at de producerede NIR-spektre fra alfalfa proteinpulveret ser lovende ud i forhold til igangsættelse af relevante yderligere undersøgelser, da en god sammenhæng mellem TDF-værdierne og NIR-spektrene kunne ses. Ved et eventuelt fremtidigt arbejde bør første prioritet være at validere denne model. Hvis det ser lovende ud, kræves både en uafhængig validering og herefter et større datasæt for at producere en ny kalibreringsmodel med en øget modelrobusthed. (Less)
Popular Abstract
Investigation of the Fibre Content in the Promising New Food Ingredient, Alfalfa Protein Powder

Sustainable food is a key topic becoming more important with time. Trends within the food sector are pointing in directions of locally grown plant based food solutions, but consumers do not want to compromise regarding nutritional qualities. In Denmark, the climate is perfectly suited for cultivation of alfalfa, a legume also known by lucerne and Medicago sativa. For centuries harvested alfalfa has been used as feed for cows being able to digest a large amount of fibres within alfalfa. Alfalfa does however also contain greater amounts of essential amino acids than for example the popular food ingredient soy, making alfalfa an interesting... (More)
Investigation of the Fibre Content in the Promising New Food Ingredient, Alfalfa Protein Powder

Sustainable food is a key topic becoming more important with time. Trends within the food sector are pointing in directions of locally grown plant based food solutions, but consumers do not want to compromise regarding nutritional qualities. In Denmark, the climate is perfectly suited for cultivation of alfalfa, a legume also known by lucerne and Medicago sativa. For centuries harvested alfalfa has been used as feed for cows being able to digest a large amount of fibres within alfalfa. Alfalfa does however also contain greater amounts of essential amino acids than for example the popular food ingredient soy, making alfalfa an interesting subject for research within the area of potential human food resources. In order to make the best use of the wanted and advantageous amino acids within the fibrous alfalfa, as much of the protein content as possible needs to be extracted from the legume. This is done by pressing harvested and wetted stems, leaves and flowers of alfalfa into a protein rich green juice, and a fibrous pulp, see Figure 1. The pH of this green juice is decreased to lower the water solubility of the wanted proteins, and thus precipitate them. The green juice is then centrifuged resulting in a pellet containing proteins amongst other alfalfa compounds. This pellet is freeze dried into a protein powder in order to concentrate the wanted proteins. This process is performed 10 times in total, the first time untreated raw wetted alfalfa is pressed into green juice as mentioned, the following nine times the fibrous pulp from the prior press is wetted and pressed into more green juice. The aim of re-pressing the fibrous pulp is to extract the highest total amount of protein from one batch of alfalfa. This protein powder
production from raw untreated alfalfa to protein powder, does increase human digestibility of alfalfa by increasing the amount of protein per weight.

The produced protein powder might have potential to be used as a food ingredient on the market when its compound composition has been further outlined. The compound composition also has to be outlined in order to extract and use the protein from alfalfa in the most cost efficient and sustainable way. Due to the composition of raw alfalfa, dietary fibres are suspected to be found in high amounts in the protein powder. Since dietary fibre determination with traditional chemical analysing methods is very time consuming, this project investigates the fibre fractions of alfalfa protein powder, and a
potential method for rapid determination of the fibre content.

In this project, the possibility to develop a model for determination of the amount and type of fibre fractions within the protein powder produced from alfalfa in a fast and cheap manner without performing traditional chemical experiments, is thus investigated. The model will be developed from near infrared (NIR) spectra of the protein powder related to enzymatically determined nutrient contents of the protein powder. NIR spectra are fingerprints of a given food sample, representing the unique physical and chemical composition of it, because the measured spectra reflect the amount of certain molecular bonds in various types of molecules, such as fibre molecules. Since NIR spectra are affected by all compounds of an analysed sample, protein, fibre, carbohydrate and ash contents are determined for each of the 10 presses leading to produced protein powder. These results are also used with the purpose of outlining the protein powder contents in general. The NIR spectra of the protein powder are preprocessed in order to produce the best correlation between the spectra and the measured fibre contents. When the best suited preprocessing methods are found, a model being able to predict the fibre content in the 10 different samples was successfully developed. The model is not validated, and therefore it is challenging to draw a conclusion regarding the model quality. For higher chances of success, and in order to produce a more robust model, big datasets, and independent validation sets are required. The results of this project do, however, encourage further investigation
and optimisation of this kind of model development. (Less)
Please use this url to cite or link to this publication:
author
Andersen, Christina LU
supervisor
organization
alternative title
Utveckling av en NIR spektroskopi-modell för prediktion av fiberföreningar i Alfalfa
course
KMBM05 20201
year
type
H2 - Master's Degree (Two Years)
subject
keywords
alfalfa, lucerne, medicago sativa, green protein powder, NIR, near infrared spectroscopy, model development, matlab, PLS, partial least squares, PCA, principal component analysis, SNV, standard normal variate, fibre prediction, TDF, total dietary fibre, applied microbiology, teknisk mikrobiologi
language
English
id
9027623
date added to LUP
2020-09-14 15:48:05
date last changed
2020-09-14 15:48:05
@misc{9027623,
  abstract     = {{Background: This project investigates if it is possible to develop a calibration model from near infrared (NIR) spectroscopic measurements, for determination of the amount and type of fibre fractions within protein powder produced from the legume alfalfa, without performing wet experiments. Alfalfa is also known as Medicago sativa and lucerne, but is in this project further referred to as alfalfa. Such a model would be applicable as a protein powder production process control, by scanning a small amount of sample during the production process, immediately resulting in a fibre content value. With this result, one will know when the process should be stopped by means of nutritional values. Except from fibres, alfalfa contains large amounts of nutrients, for example essential amino acids. The advantageous amino acids are
thus extracted from the fibrous alfalfa during the protein powder production process. The alfalfa protein powder is produced from stems, leaves and flowers of intact, freshly harvested alfalfa plants. The raw alfalfa was frozen during storage, then thawed and wetted prior to the first press, which is resulting in a protein rich green juice, and a fibrous pulp. The pH of the green juice is decreased to precipitate proteins. The green juice is then centrifuged resulting in a pellet consisting of the total water soluble solid content extracted from alfalfa. The pellet is freeze dried into a protein powder in order to concentrate the protein content. This process is performed 10 times in
total, the first time untreated raw wetted alfalfa is pressed into green juice as mentioned, the following nine times the fibrous pulp from the prior press is wetted and pressed into new samples of green juice. The aim of re-pressing the fibrous pulp is to extract the highest total amount of protein from one batch of alfalfa. This protein powder production from raw untreated alfalfa to protein powder, does increase human digestibility of alfalfa
by increasing the amount of protein per weight. Protein powder derived from each of the 10 presses was collected in separate fractions to determine to which extent the fibre profile is changing using an enzymatic gravimetric method. The amounts of protein, insoluble dietary fibre (IDF), soluble dietary fibre (SDF), total dietary fibre (TDF), available carbohydrates (ACH) and ash were determined, since NIR spectra are affected by all compounds of the protein powder. NIR spectra from all 10 presses are related directly to the determined TDF contents, which are used as reference values in order to calibrate a partial least squares (PLS) model that produces predicted TDF values. Attempts were also made to conduct NIR spectra earlier in the protein powder production process, from the green juice prior to centrifugation and from the pellet prior to freeze drying. A cellulose gluten powder dilution series comparable to
the 10 presses of protein powder was prepared, to test if a calibration model could be developed from NIR spectral data of powder containing cellulose as one of the main components. The cellulose gluten spectra were also compared with
protein powder spectra during spectral compound analyses.

Results: The nutrient profile determination resulted in a total decreasing amount of protein from 43.12% w/w for press 1 to 37.84% w/w for press 10. The TDF content increased from 22.80% w/w for press 1 to 47.47% w/w for press 10. ACH decreased from 5.43% w/w for press 1 to 1.10% w/w for press 10, while the amount of determined ash decreased from 8.24% w/w for press 1 to 2.70% w/w for press 10. Usable and promising NIR spectra were conducted from all measured protein and cellulose gluten powder samples. A calibration model predicting TDF contents for each of the 10 presses was developed with a wavenumber range from 6,800 cm-1 to 4,100 cm-1 and R2 = 0.98. For all 10 presses, the mean deviation from the reference TDF contents was 0.76% w/w. NIR spectra from the green juice and pellet could not be conducted with the available NIR instrument and presetting options.

Conclusions: It is challenging to convert complex NIR spectra into usable information. Since a broad wavenumber spectrum was chosen for the model development, it was easy to fit the spectra to almost any kind of reference
values, even though the spectra do not describe those reference values. It also has to be kept in mind that the model is not validated. Therefore it is hard to draw conclusions regarding the model quality. It can be concluded though, that
NIR spectra obtained from the protein powder of alfalfa look promising for further investigation, since a good correlation between the TDF amounts and NIR spectra could be seen. Of future work the first priority should be to validate this produced model. If that looks promising, both a new independent validation set and a larger data set to produce a new calibration model is required to further test the model robustness.}},
  author       = {{Andersen, Christina}},
  language     = {{eng}},
  note         = {{Student Paper}},
  title        = {{Development of a Near Infrared Spectroscopy Model for Prediction of Fibre Compounds in Alfalfa}},
  year         = {{2020}},
}