CN110361356A - A kind of near infrared spectrum Variable Selection improving wheat water content precision of prediction - Google Patents
A kind of near infrared spectrum Variable Selection improving wheat water content precision of prediction Download PDFInfo
- Publication number
- CN110361356A CN110361356A CN201910693888.4A CN201910693888A CN110361356A CN 110361356 A CN110361356 A CN 110361356A CN 201910693888 A CN201910693888 A CN 201910693888A CN 110361356 A CN110361356 A CN 110361356A
- Authority
- CN
- China
- Prior art keywords
- variable
- vector
- weight
- information vector
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 241000209140 Triticum Species 0.000 title claims abstract description 37
- 235000021307 Triticum Nutrition 0.000 title claims abstract description 37
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 21
- 238000002329 infrared spectrum Methods 0.000 title claims abstract description 20
- 239000013598 vector Substances 0.000 claims abstract description 59
- 238000000034 method Methods 0.000 claims abstract description 31
- 238000005070 sampling Methods 0.000 claims abstract description 16
- 238000010238 partial least squares regression Methods 0.000 claims abstract description 6
- 239000011159 matrix material Substances 0.000 claims abstract description 4
- 238000012795 verification Methods 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 8
- 238000002790 cross-validation Methods 0.000 claims description 6
- 239000000284 extract Substances 0.000 claims description 6
- 238000012804 iterative process Methods 0.000 claims description 5
- 238000010606 normalization Methods 0.000 claims description 5
- 238000010987 Kennard-Stone algorithm Methods 0.000 claims description 2
- 230000003595 spectral effect Effects 0.000 abstract description 9
- 239000000126 substance Substances 0.000 abstract description 6
- 238000004458 analytical method Methods 0.000 abstract description 4
- 238000004422 calculation algorithm Methods 0.000 description 8
- 238000001228 spectrum Methods 0.000 description 6
- 238000007621 cluster analysis Methods 0.000 description 5
- 235000013339 cereals Nutrition 0.000 description 3
- 238000009826 distribution Methods 0.000 description 3
- 230000002068 genetic effect Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 241000208340 Araliaceae Species 0.000 description 2
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 2
- 235000003140 Panax quinquefolius Nutrition 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 2
- 230000002860 competitive effect Effects 0.000 description 2
- 239000012141 concentrate Substances 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 235000008434 ginseng Nutrition 0.000 description 2
- 238000004611 spectroscopical analysis Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- LFQSCWFLJHTTHZ-UHFFFAOYSA-N Ethanol Chemical compound CCO LFQSCWFLJHTTHZ-UHFFFAOYSA-N 0.000 description 1
- 238000004566 IR spectroscopy Methods 0.000 description 1
- 229920002472 Starch Polymers 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 235000013405 beer Nutrition 0.000 description 1
- 235000015895 biscuits Nutrition 0.000 description 1
- 235000008429 bread Nutrition 0.000 description 1
- 235000012970 cakes Nutrition 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 235000019441 ethanol Nutrition 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000000855 fermentation Methods 0.000 description 1
- 230000004151 fermentation Effects 0.000 description 1
- 235000013305 food Nutrition 0.000 description 1
- 229910052500 inorganic mineral Inorganic materials 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 239000011707 mineral Substances 0.000 description 1
- 235000012149 noodles Nutrition 0.000 description 1
- 230000000050 nutritive effect Effects 0.000 description 1
- 102000004169 proteins and genes Human genes 0.000 description 1
- 108090000623 proteins and genes Proteins 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 235000019698 starch Nutrition 0.000 description 1
- 239000008107 starch Substances 0.000 description 1
- 239000004575 stone Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
- G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
- G01N21/3563—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light for analysing solids; Preparation of samples therefor
Landscapes
- Physics & Mathematics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention relates to a kind of near infrared spectrum Variable Selections for improving wheat water content precision of prediction, belong to agricultural analysis field.Specific implementation process is as follows: acquiring the near infrared spectrum data of wheat first, measures wheat water content chemical score content;Secondly, stochastical sampling is carried out to spectral variables space by binary matrix sampling method, two kinds of information vectors of occurrences frequency and Partial Least Squares Regression coefficient are done weighting to handle to obtain the contribution margin of each spectral variables, the small variable of contribution margin is deleted using decaying exponential function, generates the new variable space;Finally, generating new subset using weight sampling method based on the new variable space, it establishes recurrence submodel and obtains the weight of each variable in subset using the regression coefficient absolute value of model, optimized variable weight is gradually corrected, optimal variables set is obtained, wheat water content prediction model is established with this.This method compared with prior art, improves the precision of prediction and stability of model.
Description
Technical field
Present method invention belongs to agricultural analysis field, and in particular to a kind of near infrared light for improving wheat water content precision of prediction
Compose Variable Selection.
Background technique
Wheat is one of main cereal crops in China, the cereal crops that wheat is planted extensively as one kind, rich in starch,
Moisture, protein, fat, minerals and some needed by human body microelement, can make after grinds biscuit, cake,
Beer, alcohol etc. can be made in noodles, steamed bun, bread after fermentation, have good nutritive value, wheat water content content is to comment
Estimate the important indicator of wheat quality, the methods and techniques of Fast nondestructive evaluation wheat quality, for grain inspection and food processing
Etc. it is significant.
Near-infrared spectrum technique can simultaneously, quickly, it is lossless the multiple indexs of wheat are tested and analyzed, due to close red
External spectrum is mainly that the frequency multiplication of substance and sum of fundamental frequencies absorb, and signal is relatively weak, and bands of a spectrum are wider, overlapping is serious, it is therefore desirable to
Near infrared spectrum data is handled in conjunction with based on the stoechiometric process of variable selection algorithm, extracts the feature letter of sample
Breath, to realize the prediction to unknown sample chemical score.
Common Variable Selection has variable combination of sets cluster analysis method (Variable Combination both at home and abroad
Population Analysis, VCPA, referring to YongHuan Yun, WeiTing Wang, BaiChuan Deng.Using
variable combination population Analysis for variable selection in
Multivariate calibration. [J] .Analytica ChimicaActa.2015.862:14-23), iteration retain letter
Cease quantity method (Iteratively Retains Informative Variables, IRIV, referring to YongHuanYun,
WeiTing Wang,YiZeng Liang.A strategy that iteratively retains informative
variables for selectingoptimal variable subset in multivariate calibration.
[J] .Analytica Chimica Acta.2014,807:36-43), genetic algorithm (genetic algorithm, GA, ginseng
See Leardi R, Gonzalez AL, Genetic algorithms applied to feature selection inPLS
Regression:how and when to use them, Chemom Intell Lab Syst, 1998,41,195-207),
Competitive adaptive weight weight sampling analytic approach (CompetitiveAdaptive Reweighted Sampling CARS, ginseng
See HongDong Li, YiZeng Liang.Key wavelengths screening uing competitive
adaptive reweighted sampling method for multivariate calibration.[J]
.Analytica (1) Chimica Acta.2009,648: 77-84), from weight variable combine clusteranalysis (Automatic
Weighting Variable Combination PopulationAnalysis, AWVCPA, referring to Zhao Huan, official gram is Shi Xiao
Light studies [J] analytical chemistry based on the near infrared spectrum Variable Selection from weight variable combination clusteranalysis,
2018,46 (1): 136-142) and variable combination of sets cluster analysis iteration reservation information variable method (Variable Combination
Population Analysis-Iteratively Retains Informative Variables, VCPA-IRIV) etc..
In wheat water content content prediction, existing Variable Selection all Force Deletion secondary variable and contribution are less
Variable, ignore variable and combine influence to estimated performance, can there is important feature when these variables are combined
Information, when variables number is very big, some Variable Selections will lead to very high over-fitting risk, generate very high prediction
Error, so that prediction result is inaccurate, furthermore existing algorithm model is complicated, and precision of prediction is low, and model is unstable.
Summary of the invention
In view of the deficiencies of the prior art and defect, the invention proposes a kind of for improving the close of wheat water content precision of prediction
Infrared spectroscopy Variable Selection, this method is based on lesser cross validation root-mean-square error value, to Partial Least Squares Regression system
Weighting processing is normalized in several and two kinds of information vectors of occurrences frequency results, calculates the contribution of each spectral variables
Value, according to the size of contribution margin, establishes regression model, the regression coefficient absolute value based on model obtains variable weight, gradually school
Positive optimized variable weight, obtains optimal variables set, establishes prediction model with this, can be very good the precision for improving prediction model
And stability.
Specific step is as follows:
A measures the near infrared spectrum data X and wheat water content content chemistry Value Data Y of wheat sample, with Kennard-
Stone algorithm is divided into calibration set and forecast set;
B is sampled K times from the variable space by binary matrix sampling method, obtains K variable subset, each variable
Collection is all containing one group of random variable combination, and wherein K value is 1500;
C calculates the cross-verification root-mean-square error of each variable combination using Partial Least Squares, and chooses its interaction
Examine the smallest preceding σ × K variable subset of root-mean-square error as variables set, wherein σ value takes 15%;
D statistical variable is concentrated the frequency of each occurrences and is normalized, and then has obtained a variable weight
The property wanted judgment basis is known as type I information vector, the occurrences frequency values after normalized be with type I information to
Amount is the variable contribution margin under criterion;
E calculates Partial Least Squares Regression of each variable in different variable subsets in variables set described in step C
The absolute value of coefficient, and be normalized, the finally normalization to variable each in variables set in different variable subsets
Regression coefficient absolute value is summed, and variable normalizes the size of the sum of regression coefficient absolute value and the importance of variable at just
Than, and then obtain second variable importance criterion and be known as the second category information vector, each variable is in different variable subsets
Normalize regression coefficient absolute value and the variable contribution margin for the variable under using the second category information vector as criterion;
F according to the cross-verification root-mean-square error of every kind of information vector be arranged the second category information of type I information vector sum to
The weight of amount;
G calculates the tribute of each variable in variables set according to the weight of type I information vector sum the second category information vector
Offer value;
H deletes the variable small using the calculated contribution margin of step G with decaying exponential function, retains and is counted using step G
The big variable of the contribution margin of calculating obtains a new variable space R;
Variable in variable space R is repeated step B~step H and carries out Variable Selection, this process iteration n times, N by I
Value is 50, retains cross-verification root-mean-square error in an iterative process and is worth small set, is finally left L variable, and L value is 100;
J samples remaining L variable using self-service stochastical sampling method, generates mutually not exactly the same Z
Subset, Z value are all variables selection probability right having the same that 500, Z son is concentrated;
K establishes submodel with the Z subset obtained in step J, calculates the cross validation root-mean-square error of submodel, extracts
The best model of cross validation root-mean-square error the smallest 15% out;
L calculates the regression coefficient of each best model extracted in step K, obtains the regression vector of each best model,
The form that regression coefficients all in above-mentioned regression vector are converted to absolute value, obtains quadratic regression vector, all secondary returnings
Return vector to be normalized to obtain final regression vector, and sum to final regression vector, is asked according to final regression vector
Sum as a result, assigning each variable new weight;
New weight of the M based on each variable, application weighting sampling go to generate mutually not exactly the same new subset, and structure
The submodel for building new subset, in the submodel of new subset, the select probability for the variable for enabling regression coefficient absolute value bigger
It is worth bigger;
J~M step iteration is run n times by N, and N value is 50, in an iterative process that cross validation root-mean-square error value is minimum
Subset as optimal variables set, wheat water content prediction model is established with optimal variables set.
First information vector weight and the second information vector power according to above-mentioned Variable Selection, in the step F
The calculation formula of weight:
w1: the weight of type I information vector;w2: the weight of the second category information vector;RMSECV1: type I information vector
Cross-verification root-mean-square error;RMSECV2: the cross-verification root-mean-square error of the second category information vector;
The calculation formula of the contribution margin of each variable in variables set described in the step G:
Yi: the contribution margin of i-th of variable, the more big then variable of value are more important;I-th of variable is believed with the first kind
Ceasing vector is the variable contribution margin under criterion;I-th of variable using the second category information vector as criterion under
Variable contribution margin.
Compared with existing wheat water content analytical technology, it is proposed by the present invention it is a kind of improve wheat water content precision of prediction it is close red
External spectrum Variable Selection, the importance of judgment variable in such a way that two kinds of information vectors weight, it is contemplated that two kinds of letters
Influence of the vector to prediction model is ceased, the defect only with a kind of information variable as variable importance judgment basis is compensated for,
Processing is weighted to variable based on model regression coefficient simultaneously, considers that variable combines the influence to prediction result, reduces light
Variable is composed, prediction model is simplified, greatly improves the precision of prediction of model.
Detailed description of the invention
With reference to the accompanying drawing and embodiment the invention will be further described:
Fig. 1 is a kind of near infrared spectrum Variable Selection flow chart for improving wheat water content precision of prediction of the present invention
Fig. 2 is the atlas of near infrared spectra of wheat sample
Scatter plot distributions of the Fig. 3 between forecast set true value and predicted value
Fig. 4 is averaged spectrum and the final selected characteristic variable distribution map of every kind of Variable Selection
Specific embodiment
Embodiment one: it in order to prove applicability of the invention, is described in detail in conjunction with example.But the present invention
It can be applied to the spectroscopic data except this used example.
Fig. 1 is a kind of stream of near infrared spectrum Variable Selection for improving wheat water content precision of prediction provided by the invention
Cheng Tu, it is seen then that the present invention specifically includes the following steps:
(1) wheat near infrared spectrum data collected by contains 66 wheat samples, the near infrared spectrum of each sample
Wavelength distribution tests the near infrared spectrum of each wheat sample with spectrometer in 950-1700nm, and chemically tests
The chemical score of each sample moisture content content.Wherein 44 sample spectrum data are chosen with Kennard-Stone (K-S) method
Establish prediction model as calibration set with chemical Value Data, using the spectroscopic data of remaining 22 samples and chemical Value Data as
The feasibility of forecast set Sample model, wheat atlas of near infrared spectra are as shown in Figure 2.
(2) it samples 1500 times and obtains from the wheat near infrared spectrum variable space with binary matrix sampling method (BMS)
1500 groups of different variable subsets calculate the friendship of this 1500 groups of difference variable subsets with Partial Least Squares (PLS) later
Root-mean-square error (RMSECV) mutually is examined, chooses the smallest preceding 15% group of variable subset of its RMSECV value as variables set.
(3) record variable concentrate the frequency of occurrence of each spectral variables and being normalized obtain type I information to
Amount.
(4) it records Partial Least Squares Regression coefficient of each spectral variables in variables set and is normalized, most
Afterwards the absolute value of the normalization Partial Least Squares Regression coefficient of variable identical in variables set is summed to obtain the second category information
Vector.
(5) weight of these two types of information vectors is calculated separately by formula (I) (II), and change is calculated according to formula (III)
The contribution margin of each spectral variables in quantity set.
The weight calculation formula of information vector
w1: the weight of type I information vector;w2: the weight of the second category information vector;RMSECV1: type I information vector
Cross-verification root-mean-square error;RMSECV2: the cross-verification root-mean-square error of the second category information vector;
The contribution margin calculation formula of each spectral variables is as follows:
Yi: the contribution margin of i-th of variable, the more big then variable of value are more important;I-th of variable is believed with the first kind
Ceasing vector is the variable contribution margin under criterion;I-th of variable using the second category information vector as criterion under
Variable contribution margin.
(6) the small spectral variables of those contribution margins are deleted with decaying exponential function, retains the big spectrum of its contribution margin and becomes
Amount, obtains a new variable space R.
rN=e-θ×N (Ⅳ)
rN: decaying exponential function runs n times variations per hour retention rate;θ: curve controlled parameter, it and decaying exponential function
Execution number is related, and the number that decaying exponential function executes is more, and θ value is smaller.N: the execution number of decaying exponential function.It is bent
The calculation formula of line traffic control parameter is
(7) (2)~(6) process is repeated to the variable in R, this process iteration 50 times is finally left 100 spectral variables.
(8) 500 subsets are generated using self-service stochastical sampling method to remaining variable in variable space R, in every height
It concentrates and extracts selected variable, and reject repeated variable, the variable subset of acquisition is established into submodel, calculates submodel
RMSECV, and best model is extracted by small RMSECV.
(9) each regression coefficient RC for extracting model is calculated, all regression vectors are normalized between (0,1), and right
Regression vector summation, so that variable obtains new weight Wi.New weight WiCalculation formula is as follows:
Wi: new weight, Vi,Z: the normalization regression coefficient absolute value of i-th of variable in Z subset.
(10) the new weight based on variable, application weighting sampling go to generate new subset, extract institute in each new subset
The variable of selection rejects repeated variable and constructs submodel, assigns the big variable of regression coefficient absolute value to big weight.
(11) by (8)~(10) step iteration run 50 times, in an iterative process using the lesser subset of RMSECV value as
Optimal variables set.
(12) wheat water content prediction model is established with optimal variables set.
In order to avoid influence of the algorithmic theory of randomness to variables choice result in algorithm operational process, will be combined from weight variable
Cluster analysis combination weight sampling algorithm (AWVCPA-WS) is run 50 times, chooses in AWVCPA-WS highest one group of precision of prediction
Characteristic variable as final characteristic variable choose as a result, establishing moisture content in wheat eventually by AWVCPA-WS-PLS
The prediction result of prediction model is as shown in Figure 3.
In order to illustrate the superiority of AWVCPA-WS Variable Selection, by wheat near infrared spectrum data in identical item
Six kinds of Variable Selections of GA, IRIV, VCPA, CARS, VCPA-IRIV and AWVCPA have been respectively adopted under part and have carried out characteristic variable
It extracts, since every kind of Variable Selection all has certain randomness in the process of running, and then influences the reliability of model,
So we run above every kind of Variable Selection 50 times, it is pre- in modeling process to calculate every kind of Variable Selection
Root-mean-square error average value is surveyed, and selects the highest one group of characteristic variable of its every kind algorithm precision of prediction as final characteristic variable
It chooses as a result, establishing prediction model using PLS, characteristic variable result selected by every kind of Variable Selection is as shown in figure 4, every
Kind modeling method the results are shown in Table 1.
Prediction result of the different modeling methods of table 1 to wheat water content content
Embodiment of the present invention explanation leaves it at that.
Claims (2)
1. a kind of near infrared spectrum Variable Selection for improving wheat water content precision of prediction, which is characterized in that include following step
It is rapid:
A measures the near infrared spectrum data X and wheat water content content chemistry Value Data Y of wheat sample, with Kennard-Stone
Algorithm is divided into calibration set and forecast set;
B is sampled K times from the variable space by binary matrix sampling method, obtains K variable subset, each variable subset
Containing one group of random variable combination, wherein K value is 1500;
C calculates the cross-verification root-mean-square error of each variable combination using Partial Least Squares, and chooses its cross-verification
The smallest preceding σ × K variable subset of root-mean-square error is as variables set, and wherein σ value takes 15%;
D statistical variable is concentrated the frequency of each occurrences and is normalized, and then has obtained a variable importance
Judgment basis is known as type I information vector, and the occurrences frequency values after normalized are to be with type I information vector
Variable contribution margin under criterion;
E calculates Partial Least Squares Regression coefficient of each variable in different variable subsets in variables set described in step C
Absolute value, and be normalized, finally the normalization to variable each in variables set in different variable subsets returns
Absolute coefficient is summed, and the size that variable normalizes the sum of regression coefficient absolute value is directly proportional to the importance of variable, into
And it obtains second variable importance criterion and is known as the second category information vector, normalization of each variable in different variable subsets
Regression coefficient absolute value and variable contribution margin for the variable under using the second category information vector as criterion;
The second category information of type I information vector sum vector is arranged according to the cross-verification root-mean-square error of every kind of information vector in F
Weight;
G calculates the contribution margin of each variable in variables set according to the weight of type I information vector sum the second category information vector;
H deletes the variable small using the calculated contribution margin of step G with decaying exponential function, and reservation is calculated using step G
The big variable of contribution margin, obtain a new variable space R;
Variable in variable space R is repeated step B~step H and carries out Variable Selection, this process iteration n times by I, and N value is
50, retain cross-verification root-mean-square error in an iterative process and be worth small set, be finally left L variable, L value is 100;
J samples remaining L variable using self-service stochastical sampling method, generates Z mutually not exactly the same subset,
Z value is all variables selection probability right having the same that 500, Z son is concentrated;
K establishes submodel with the Z subset obtained in step J, calculates the cross validation root-mean-square error of submodel, extracts friendship
The best model of fork verifying root-mean-square error the smallest 15%;
L calculates the regression coefficient of each best model extracted in step K, obtains the regression vector of each best model, will be upper
The form that all regression coefficients in regression vector are converted to absolute value is stated, quadratic regression vector is obtained, all quadratic regressions are sweared
Amount is normalized to obtain final regression vector, and sums to final regression vector, according to the summation of final regression vector
As a result, assigning each variable new weight;
New weight of the M based on each variable, application weighting sampling goes to generate mutually not exactly the same new subset, and constructs new
Subset submodel, in the submodel of new subset, the select probability value for the variable for enabling regression coefficient absolute value bigger is more
Greatly;
J~M step iteration is run n times by N, and N value is 50, and cross validation root-mean-square error is worth the smallest son in an iterative process
Collection is used as optimal variables set, establishes wheat water content prediction model with optimal variables set.
2. a kind of near infrared spectrum Variable Selection for improving wheat water content precision of prediction described according to claim 1,
It is characterized in that, the calculation formula of first information vector weight and the second information vector weight in the step F:
w1: the weight of type I information vector;w2: the weight of the second category information vector;RMSECV1: the friendship of type I information vector
Mutually examine root-mean-square error;RMSECV2: the cross-verification root-mean-square error of the second category information vector;
The calculation formula of the contribution margin of each variable in variables set described in the step G:
Yi: the contribution margin of i-th of variable, the more big then variable of value are more important;I-th of variable with type I information to
Amount is the variable contribution margin under criterion;I-th of variable using the second category information vector as criterion under variable
Contribution margin.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910693888.4A CN110361356A (en) | 2019-07-30 | 2019-07-30 | A kind of near infrared spectrum Variable Selection improving wheat water content precision of prediction |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910693888.4A CN110361356A (en) | 2019-07-30 | 2019-07-30 | A kind of near infrared spectrum Variable Selection improving wheat water content precision of prediction |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110361356A true CN110361356A (en) | 2019-10-22 |
Family
ID=68222801
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910693888.4A Pending CN110361356A (en) | 2019-07-30 | 2019-07-30 | A kind of near infrared spectrum Variable Selection improving wheat water content precision of prediction |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110361356A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110907393A (en) * | 2019-11-22 | 2020-03-24 | 黑龙江八一农垦大学 | Method and device for detecting saline-alkali stress degree of plants |
CN111474136A (en) * | 2020-04-26 | 2020-07-31 | 长春理工大学 | Near infrared spectrum variable selection method for improving prediction accuracy of oil content in corn |
CN111504942A (en) * | 2020-04-26 | 2020-08-07 | 长春理工大学 | Near infrared spectrum analysis method for improving prediction accuracy of protein in milk |
CN111999258A (en) * | 2020-07-03 | 2020-11-27 | 桂林理工大学 | Spectral baseline correction-oriented weighting modeling local optimization method |
CN114354666A (en) * | 2021-12-31 | 2022-04-15 | 三峡大学 | Method for extracting and optimizing spectral characteristics of soil heavy metal based on wavelength frequency selection |
CN117805024A (en) * | 2024-02-29 | 2024-04-02 | 四川农业大学 | Crisp pear saccharin degree detection method and device, cloud equipment and computer device |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2015179985A1 (en) * | 2014-05-29 | 2015-12-03 | Canadian Light Source Inc. | Screening method for selection of disease-resistant plant cultivars |
CN106950193A (en) * | 2017-05-24 | 2017-07-14 | 长春理工大学 | Based on the near infrared spectrum Variable Selection that cluster analysis is combined from weight variable |
-
2019
- 2019-07-30 CN CN201910693888.4A patent/CN110361356A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2015179985A1 (en) * | 2014-05-29 | 2015-12-03 | Canadian Light Source Inc. | Screening method for selection of disease-resistant plant cultivars |
CN106950193A (en) * | 2017-05-24 | 2017-07-14 | 长春理工大学 | Based on the near infrared spectrum Variable Selection that cluster analysis is combined from weight variable |
Non-Patent Citations (2)
Title |
---|
宋雨宸: "基于模型集群分析的近红外光谱变量选择方法研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
赵环 等: "基于自加权变量组合集群分析法的近红外光谱变量选择方法研究", 《分析化学(FENXI HUAXUE)研究报告》 * |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110907393A (en) * | 2019-11-22 | 2020-03-24 | 黑龙江八一农垦大学 | Method and device for detecting saline-alkali stress degree of plants |
CN111474136A (en) * | 2020-04-26 | 2020-07-31 | 长春理工大学 | Near infrared spectrum variable selection method for improving prediction accuracy of oil content in corn |
CN111504942A (en) * | 2020-04-26 | 2020-08-07 | 长春理工大学 | Near infrared spectrum analysis method for improving prediction accuracy of protein in milk |
CN111999258A (en) * | 2020-07-03 | 2020-11-27 | 桂林理工大学 | Spectral baseline correction-oriented weighting modeling local optimization method |
CN111999258B (en) * | 2020-07-03 | 2022-06-17 | 桂林理工大学 | Spectral baseline correction-oriented weighting modeling local optimization method |
CN114354666A (en) * | 2021-12-31 | 2022-04-15 | 三峡大学 | Method for extracting and optimizing spectral characteristics of soil heavy metal based on wavelength frequency selection |
CN114354666B (en) * | 2021-12-31 | 2023-12-26 | 三峡大学 | Soil heavy metal spectral feature extraction and optimization method based on wavelength frequency selection |
CN117805024A (en) * | 2024-02-29 | 2024-04-02 | 四川农业大学 | Crisp pear saccharin degree detection method and device, cloud equipment and computer device |
CN117805024B (en) * | 2024-02-29 | 2024-05-14 | 四川农业大学 | Crisp pear saccharin degree detection method and device, cloud equipment and computer device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110361356A (en) | A kind of near infrared spectrum Variable Selection improving wheat water content precision of prediction | |
Theanjumpol et al. | Non-destructive identification and estimation of granulation in ‘Sai Num Pung’tangerine fruit using near infrared spectroscopy and chemometrics | |
Baca-Bocanegra et al. | Feasibility study on the use of a portable micro near infrared spectroscopy device for the “in vineyard” screening of extractable polyphenols in red grape skins | |
CN111504942A (en) | Near infrared spectrum analysis method for improving prediction accuracy of protein in milk | |
CN101738373A (en) | Method for distinguishing varieties of crop seeds | |
CN106841083A (en) | Sesame oil quality detecting method based on near-infrared spectrum technique | |
CN105044024A (en) | Method for nondestructive testing of grape berries based on near infrared spectrum technology | |
Fadock et al. | Visible-near infrared reflectance spectroscopy for nondestructive analysis of red wine grapes | |
CN106951720A (en) | Soil nutrient model transfer method based on canonical correlation analysis and linear interpolation | |
Saad et al. | Quality analysis prediction and discriminating strawberry maturity with a hand-held Vis–NIR spectrometer | |
CN112730312A (en) | Doped bovine colostrum qualitative identification method based on near infrared spectrum technology | |
CN108613943A (en) | A kind of near-infrared single grain crop component detection method based on the transfer of spectrum form | |
CN115201148A (en) | Method for rapidly detecting content of fatty acid in rice based on near infrared spectrum | |
CN108344701A (en) | Paraffin grade qualitative classification based on hyperspectral technique and quantitative homing method | |
CN110632017A (en) | Rapid identification optimization model for transgenic soybeans based on morphological analysis | |
CN110231305A (en) | A method of DPPH free radical scavenging ability in the odd sub- seed of measurement | |
Du et al. | Quantitative detection of azodicarbonamide in wheat flour by near-infrared spectroscopy based on two-step feature selection | |
CN106950193B (en) | Based on the near infrared spectrum Variable Selection from weight variable combination cluster analysis | |
CN113310934A (en) | Method for quickly identifying milk cow milk mixed in camel milk and mixing proportion thereof | |
Zhang et al. | Hyperspectral model based on genetic algorithm and SA-1DCNN for predicting Chinese cabbage chlorophyll content | |
Wang et al. | Monitoring model for predicting maize grain moisture at the filling stage using NIRS and a small sample size | |
CN111624317A (en) | Nondestructive testing method for judging freshness of baby cabbage | |
US20230089466A1 (en) | Establishment of Identification and Screening Method of Cows with A2 Beta-Casein Genotype of Producing A2 Milk and Applications Thereof | |
Yuan et al. | Application of hyperspectral imaging to discriminate waxy corn seed vigour after aging. | |
CN115630332A (en) | Wheat flour quality characteristic prediction method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191022 |