CN100444153C - Cigarette internal quality index extimating method based on regression function estimating SVM - Google Patents
Cigarette internal quality index extimating method based on regression function estimating SVM Download PDFInfo
- Publication number
- CN100444153C CN100444153C CNB2005101052856A CN200510105285A CN100444153C CN 100444153 C CN100444153 C CN 100444153C CN B2005101052856 A CNB2005101052856 A CN B2005101052856A CN 200510105285 A CN200510105285 A CN 200510105285A CN 100444153 C CN100444153 C CN 100444153C
- Authority
- CN
- China
- Prior art keywords
- cigarette
- svm
- regression function
- evaluation
- svm model
- 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.)
- Expired - Fee Related
Links
- 235000019504 cigarettes Nutrition 0.000 title claims abstract description 56
- 238000000034 method Methods 0.000 title claims abstract description 44
- 230000006870 function Effects 0.000 claims description 52
- 238000012706 support-vector machine Methods 0.000 claims description 48
- 238000011156 evaluation Methods 0.000 claims description 32
- 230000001953 sensory effect Effects 0.000 claims description 16
- 239000000779 smoke Substances 0.000 claims description 13
- 239000013598 vector Substances 0.000 claims description 11
- 238000010219 correlation analysis Methods 0.000 claims description 9
- 230000000391 smoking effect Effects 0.000 claims description 7
- 238000004458 analytical method Methods 0.000 claims description 6
- 238000000513 principal component analysis Methods 0.000 claims description 5
- 238000006243 chemical reaction Methods 0.000 claims description 4
- 238000012545 processing Methods 0.000 claims description 4
- 238000010606 normalization Methods 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 238000012549 training Methods 0.000 claims description 3
- FGUUSXIOTUKUDN-IBGZPJMESA-N C1(=CC=CC=C1)N1C2=C(NC([C@H](C1)NC=1OC(=NN=1)C1=CC=CC=C1)=O)C=CC=C2 Chemical compound C1(=CC=CC=C1)N1C2=C(NC([C@H](C1)NC=1OC(=NN=1)C1=CC=CC=C1)=O)C=CC=C2 FGUUSXIOTUKUDN-IBGZPJMESA-N 0.000 claims 1
- 238000002790 cross-validation Methods 0.000 claims 1
- 238000001303 quality assessment method Methods 0.000 abstract description 4
- 241000208125 Nicotiana Species 0.000 description 11
- 235000002637 Nicotiana tabacum Nutrition 0.000 description 11
- 238000013441 quality evaluation Methods 0.000 description 8
- 239000000126 substance Substances 0.000 description 4
- 230000007547 defect Effects 0.000 description 3
- 238000013461 design Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- SNICXCGAKADSCV-JTQLQIEISA-N (-)-Nicotine Chemical compound CN1CCC[C@H]1C1=CC=CN=C1 SNICXCGAKADSCV-JTQLQIEISA-N 0.000 description 2
- IJGRMHOSHXDMSA-UHFFFAOYSA-N Atomic nitrogen Chemical compound N#N IJGRMHOSHXDMSA-UHFFFAOYSA-N 0.000 description 2
- 206010013911 Dysgeusia Diseases 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 239000003205 fragrance Substances 0.000 description 2
- 230000007794 irritation Effects 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 229960002715 nicotine Drugs 0.000 description 2
- SNICXCGAKADSCV-UHFFFAOYSA-N nicotine Natural products CN1CCCC1C1=CC=CN=C1 SNICXCGAKADSCV-UHFFFAOYSA-N 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- ZAMOUSCENKQFHK-UHFFFAOYSA-N Chlorine atom Chemical compound [Cl] ZAMOUSCENKQFHK-UHFFFAOYSA-N 0.000 description 1
- 206010035148 Plague Diseases 0.000 description 1
- ZLMJMSJWJFRBEC-UHFFFAOYSA-N Potassium Chemical compound [K] ZLMJMSJWJFRBEC-UHFFFAOYSA-N 0.000 description 1
- 241000607479 Yersinia pestis Species 0.000 description 1
- NMLQNVRHVSWEGS-UHFFFAOYSA-N [Cl].[K] Chemical compound [Cl].[K] NMLQNVRHVSWEGS-UHFFFAOYSA-N 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 239000000460 chlorine Substances 0.000 description 1
- 229910052801 chlorine Inorganic materials 0.000 description 1
- 239000000470 constituent Substances 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013210 evaluation model Methods 0.000 description 1
- 239000004615 ingredient Substances 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 229910052757 nitrogen Inorganic materials 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 239000011591 potassium Substances 0.000 description 1
- 229910052700 potassium Inorganic materials 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 102000004169 proteins and genes Human genes 0.000 description 1
- 108090000623 proteins and genes Proteins 0.000 description 1
- 238000004445 quantitative analysis Methods 0.000 description 1
- 230000035909 sensory irritation Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a cigarette inherent quality assessment method based on the regression function estimation SVM. The method is to set up SVM model that use regression function estimation based on the cigarette sample data, then assess the cigarette inherent qualities with the mentioned SVM model. The invention utilizes dependency analyzing method to select characterized options.
Description
Technical Field
The invention relates to a support vector machine method for more accurately realizing evaluation and prediction of cigarette internal quality indexes under the condition of limited samples, in particular to an evaluation method of the cigarette internal quality indexes by utilizing an SVM (support vector machine) model estimated by a regression function.
Background
The inherent quality of cigarettes, such as sensory quality, is closely related to certain physicochemical components in cigarettes, which determine the inherent quality of cigarettes to some extent, and nicotine in tobacco, for example, is related to sensory irritation and strength characteristics. Domain experts often need to refer to these detected ingredients when performing sensory evaluations.
Tobacco leaves contain numerous chemical constituents; the interaction of the chemical components in the smoking process stimulates the taste, smell and touch of people, and is extremely complex. The experience of the tobacco leaf evaluation expert is very expensive, but the method also has the defects of obvious uncertainty, subjectivity and the like. The tobacco industry has accumulated a lot of valuable experiences and data in the aspects of tobacco leaf chemical composition, sensory evaluation and smoke analysis, and the valuable experiences and data are not well utilized. Enterprises hope to utilize the accumulated sample data to realize computer-aided quality assessment by modeling the learning problem, so that the internal quality assessment of unknown samples is realized, the dependence on experts in the aspect of internal quality assessment is reduced, the production design cost of the enterprises is reduced, and the product quality and the management level are improved.
The support vector machine method is an intelligent technology, is suitable for solving the complex problems of nondeterministic property and difficult to solve by using the traditional mathematical model method, and is characterized by the processing capacity of small sample data; the method is good at generalizing and extracting the knowledge close to the real rule of the sample from a small amount of sample data.
The support vector machine is a nonparametric machine learning method, and well realizes the design idea of the Structure Risk Minimization (SRM) principle; many problems which plague the past machine learning method, such as the model selection problem in the neural network, the over-learning and under-learning problems, the nonlinear and dimensionality disaster problem, the local minimum point and the like, are well solved in the SVM. The basic idea behind the support vector machine is to map the input vector X to a high-dimensional feature space Z by some kind of pre-selected non-linear mapping and then construct the optimal classification hyperplane in this feature space. The above non-linear mapping is achieved by defining an appropriate inner product function.
At present, the tobacco industry mostly adopts the traditional statistical analysis method, such as multiple regression, for the quantitative analysis method among the chemical components of tobacco, the smoke analysis index and the sensory quality evaluation index. The model constructed by the traditional method only has certain applicability to the analyzed tobacco sample data; once a new sample exists, the evaluation model is often reconstructed, the solving process is complex, and quick evaluation modeling and computer-aided quality evaluation are difficult to realize.
Disclosure of Invention
In order to overcome the defects in the prior art, the invention provides a method for evaluating the internal quality index of a cigarette based on a regression function estimation SVM, which combines the technologies and methods of correlation analysis, regression function estimation SVM, evaluation output value conversion rule and the like, and correspondingly improves the algorithm and key parameters to make the method more suitable for the data environment of the cigarette industry; the method mainly aims to establish a machine learning model suitable for a small sample data environment in the cigarette industry, and fully adapts to four important data characteristics of a small sample, high dimension, large noise and nonlinearity; therefore, the demand on sample data is reduced, the manual smoking amount is reduced, the detection cost of enterprises is reduced, higher evaluation prediction conformity can be achieved under the condition of effective samples, and the requirement of computer-aided quality evaluation in the tobacco industry is met; meanwhile, a good precondition technical method is provided for the intelligent design of the cigarette leaf group formula.
In order to achieve the purpose, the invention provides an evaluation method of cigarette intrinsic quality indexes, which is characterized in that an SVM model estimated by using a regression function is established on the basis of sample data of cigarettes, and then the intrinsic quality of the cigarettes to be measured is evaluated by using the SVM model.
The SVM model for establishing the regression function estimation is realized by the following steps:
1) firstly, recording physicochemical indexes, smoke analysis indexes and smoking evaluation results of various cigarettes into a database;
2) then, judging and eliminating errors or specific samples by utilizing a zero value/null value or exceeding the maximum and minimum value range, and carrying out standardization or normalization processing on the sample data;
3) grouping and naming the last group of obtained data according to different cigarette types or physicochemical characteristics, performing characteristic parameter selection on each quality index to be evaluated by using a correlation coefficient solving or principal component analysis correlation analysis method, reserving main influence parameters, and removing secondary influence parameters, thereby determining physicochemical index input parameter items of each index to be evaluated and predicted;
4) selecting a Gaussian function or a polynomial function as a kernel function of the support vector machine; an insensitive loss function epsilon is adopted, and a cross verification method or a leave-one method is utilized to determine a kernel function parameter gamma and a regularization parameter item C of the support vector machine;
5) sending the standardized sample data into a regression function estimation support vector machine for learning and training to obtain support vectors and coefficients alpha i of the support vectors, so as to form an SVM model;
6) and storing various parameter information, support vector information and coefficients of the SVM model built in the last step into a database.
The evaluation of the internal quality of the cigarette to be tested is realized by the following steps:
1) inputting sample data of the cigarette to be detected;
2) standardizing or normalizing the sample data;
3) screening each item of data obtained in the step 2) by using a correlation coefficient solving or principal component analysis correlation analysis method, and reserving main influence parameters;
4) judging a sample group to which the data obtained in the last step belongs according to the type of the cigarette, and reading the name of the SVM model corresponding to the type of the cigarette according to the judgment result;
5) inputting the data obtained in the step 3) into the SVM model to obtain a sensory evaluation and a smoke index prediction value;
also can be provided with the step 6): and converting the sensory evaluation and smoke index prediction value obtained in the step 5) into an expression mode which can be understood by the user according to a conversion rule and outputting the expression mode.
And (3) evaluating the accuracy of the predicted value obtained in the step 4) by taking whether the ratio of the error of the output real value to the error of the target value is smaller than the allowable error as a standard.
The method mainly comprises the steps of selecting characteristic parameters by using a correlation analysis technology, learning and modeling cigarette internal quality evaluation data by using an improved regression function estimation SVM, evaluating the cigarette internal quality by using a constructed discrimination function, and finally converting an evaluation output value into an expression mode which can be understood by a user according to a certain defined rule, so that a set of complete data analysis and processing method for cigarette internal quality evaluation prediction is formed, and the problem of small samples in computer-assisted cigarette quality evaluation is effectively solved; the problem of evaluating the internal quality index of the cigarette under the condition of small sample data in the tobacco industry is solved; the demand on sample data is reduced, the manual smoking amount is reduced, and the detection cost of enterprises is reduced.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a flowchart of an SVM model for regression function estimation according to the present invention;
Detailed Description
The invention is described below with reference to the accompanying drawings and the detailed description. The invention provides a method for evaluating the internal quality index of a cigarette based on a regression function estimation SVM, which is characterized in that an SVM model estimated by the regression function is established on the basis of sample data of the cigarette, and then the internal quality of the cigarette to be tested is evaluated by the SVM model.
The invention firstly utilizes a correlation analysis method comprising a correlation coefficient solving method and an SVML method to select characteristic parameters and reduce the number of input parameters, thereby further reducing the demand of sample data during learning. In the aspect of solving the small sample learning problem, the support vector machine has more outstanding advantages than other technical methods. However, in the cigarette quality evaluation and prediction, the internal quality indexes of the cigarette, such as sensory quality including odor type, aroma quality, strength, combustibility, irritation and the like, are mostly represented by category modes, and the industry experts describe the sensory quality of the product by different characters and also give quantitative characteristics of the sensory quality in a digital form, such as the odor type and aftertaste indexes in the cigarette quality indexes are represented as follows:
odor index: the faint scent is 1, the middle of the faint scent is 2, the middle of the faint scent is 3, the middle fragrance is 4, the middle of the faint scent is 5, the middle of the heavy scent is 6, the heavy scent is 7, and the specific fragrance is 8.
Aftertaste index: comfort is "5", more comfort is "4", more comfort is "3", less comfort is "2", and the difference is "1".
Thus, the learning problem of sensory evaluation can be treated as one "multi-class partition" when solving the learning problem. However, because the traditional SVM has obvious defects in multi-class identification, the invention provides a method for modeling by utilizing a regression function estimation SVM to solve the multi-class pattern identification problem of sensory evaluation, and the estimation of the regression function is realized by constructing a learning model with excellent popularization performance, good anti-noise performance and robustness on sample data, so that the quality evaluation of a newly added sample is completed by applying the estimated regression function.
Therefore, the advantages of the support vector machine in the learning of small samples can be exerted, and the problem of multi-class identification in the evaluation and prediction of the internal quality indexes of the cigarettes can be solved.
As shown in fig. 1, the SVM model for establishing regression function estimation is implemented by the following steps:
1) firstly, recording physicochemical indexes, smoke analysis indexes and smoking evaluation results of various cigarettes into a database;
2) then, judging and eliminating errors or specific samples by utilizing a zero value/null value or exceeding the maximum and minimum value range, and carrying out standardization or normalization processing on the sample data;
3) grouping and naming the last group of obtained data according to different cigarette types or physicochemical characteristics, performing characteristic parameter selection on each quality index to be evaluated by using a correlation coefficient solving or principal component analysis correlation analysis method, reserving main influence parameters, and removing secondary influence parameters, thereby determining physicochemical index input parameter items of each index to be evaluated and predicted; for example, the final analysis yields 10 physicochemical indices of the single-dose cigarettes: taking total sugar, total nicotine, reducing sugar, total nitrogen, protein, chlorine, potassium, Schenk value, sugar-base ratio and potassium-chlorine ratio as input parameters of quality indexes such as odor type, irritation, CO, TPM and the like;
4) selecting a Gaussian function or a polynomial function as a kernel function of the support vector machine; an insensitive loss function epsilon is adopted, and a cross verification method or a leave-one method is utilized to determine a kernel function parameter gamma and a regularization parameter item C of the support vector machine;
5) inputting the standardized sample data into a regression function estimation support vector machine for learning and training to obtain the support vector and the coefficient alpha of each support vectoriThereby forming an SVM model;
6) and storing various parameter information, support vector information and coefficients of the SVM model built in the last step into a database.
In the process of establishing the SVM model, an insensitive loss function epsilon is adopted to ensure the stability of the SVM. In order to construct a support vector machine for a real-valued function, a new loss function, namely an epsilon-insensitive loss function, is adopted in the invention:
L(y，f(x，w))＝L(|y-f(x，w)|ε)
wherein,
the epsilon insensitive loss function is an extension of the minimum maximum theoretical mean absolute error criterion proposed by Huber, and the algorithm is robust. The minimum maximum theory proposed by Huber; so that we can find the best strategy for the loss function with knowledge of only general information about the noise model.
As shown in fig. 2, the method for evaluating the internal quality index of the cigarette specifically comprises the following steps: firstly, collecting data such as physicochemical indexes, smoke indexes and smoking evaluation results of cigarettes to be evaluated, and inputting the data into a database; and standardizing or normalizing the sample data; then screening is carried out according to the correlation analysis conclusion, and wrong or specific samples are removed so as to unify all input data dimensions; judging a sample group to which the data obtained in the last step belongs according to the type of the cigarette, reading an SVM model name corresponding to the type of the cigarette according to a judgment result, and customizing the adopted SVM model by a user; then reading a kernel function, a regularization parameter C, a kernel function parameter gamma and a coefficient alpha of each support vector of the model according to the name of the model, constructing a discrimination function according to the parameters, sending data into the SVM model, and finally calculating to obtain a sensory evaluation and smoke index predicted value; the accuracy evaluation of the obtained predicted value takes as a criterion whether the ratio of the error of the output real value to the target value is smaller than the allowable error. The regression function obtained by the SVM method provided by the invention is a function approximation of a real-value sample data set, so that the output of the regression function is also in a real number form; the real number form reflects the membership degree of the estimated value and the actual class to a certain extent, and can map the proximity degree of the sample points in the input sample space into the output value, so that the real condition is closer to the real condition. However, how to evaluate the accuracy of the regression function output value becomes a new problem. Aiming at the problem, the invention provides a corresponding solution; the ratio of the error of the output real value to the target value and the allowable error is used as the calculation standard for evaluating the accuracy, as follows:
that is, if the error between the real value of the SVM model output and the target value is larger than the allowable error, the accuracy of the evaluation of the sample is 0. The performance condition of the established SVM model can be effectively estimated through the calculation standard.
Therefore, the advantages of the support vector machine in the learning of small samples can be exerted, and the problem of multi-class identification in the evaluation and prediction of the internal quality indexes of the cigarettes can be solved.
And converting the sensory evaluation and smoke index prediction value obtained in the last step into data or character type description forms which can be understood by the user according to the conversion rule and outputting the data or character type description forms for the user to refer.
Practice proves that the method is an effective intelligent method for evaluating and predicting the internal quality of single-material cigarettes or finished product cigarettes in the tobacco industry, such as sensory evaluation, smoke indexes and the like, and has more outstanding advantages in a small sample data environment compared with the existing method.
Claims (3)
1. A method for evaluating the internal quality index of cigarette based on regression function to estimate SVM is characterized by that on the basis of the sample data of cigarette, an SVM model estimated by regression function is built, then the above-mentioned SVM model is used to evaluate the internal quality of cigarette to be measured,
the SVM model for establishing the regression function estimation is realized by the following steps:
1) firstly, recording physicochemical indexes, smoke analysis indexes and smoking evaluation results of various cigarettes into a database;
2) then, judging and eliminating errors or specific samples by utilizing a zero value/null value or exceeding the maximum and minimum value range, and carrying out standardization or normalization processing on the sample data;
3) grouping and naming the data obtained in the last step according to the difference of cigarette types or physicochemical characteristics, selecting characteristic parameters of each quality index to be evaluated by using a correlation coefficient solving or principal component analysis correlation analysis method, reserving main influence parameters, and removing secondary influence parameters, thereby determining physicochemical index input parameter items of each index to be evaluated and predicted;
4) selecting a Gaussian function or a polynomial function as a kernel function of the support vector machine; determining a kernel function parameter gamma and a regularization parameter item C of the support vector machine by adopting an insensitive loss function epsilon and utilizing a cross validation method or a leave-one-out method;
5) sending the standardized sample data into a regression function estimation support vector machine for learning and training to obtain support vectors and coefficients alpha i of the support vectors, so as to form an SVM model;
6) storing each item of parameter information, support vector information and each coefficient of the SVM model built in the last step into a database,
the evaluation of the internal quality of the cigarette to be tested is realized by the following steps:
a) inputting sample data of the cigarette to be detected;
b) standardizing or normalizing the sample data;
c) screening each item of data obtained in the step b) by using a correlation coefficient solving or principal component analysis correlation analysis method, and reserving main influence parameters;
d) judging a sample group to which the data obtained in the last step belongs according to the type of the cigarette, and reading the name of the SVM model corresponding to the type of the cigarette according to the judgment result;
e) inputting the data obtained in the step c) into the SVM model, and obtaining a sensory evaluation and a smoke index prediction value.
2. The regression function estimation SVM based cigarette internal quality index evaluation method according to claim 1, wherein: further comprising step f): and e) converting the sensory evaluation and smoke index prediction value obtained in the step e) into an expression mode which can be understood by a user according to a conversion rule and outputting the expression mode.
3. The method for evaluating intrinsic quality indexes of cigarettes based on regression function estimation SVM according to claim 1, wherein the accuracy evaluation of the predicted value obtained in step e) uses as a criterion whether the ratio of the error of the output real value to the error of the target value is less than the allowable error.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005101052856A CN100444153C (en) | 2005-11-28 | 2005-11-28 | Cigarette internal quality index extimating method based on regression function estimating SVM |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005101052856A CN100444153C (en) | 2005-11-28 | 2005-11-28 | Cigarette internal quality index extimating method based on regression function estimating SVM |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1975705A CN1975705A (en) | 2007-06-06 |
| CN100444153C true CN100444153C (en) | 2008-12-17 |
Family
ID=38125779
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2005101052856A Expired - Fee Related CN100444153C (en) | 2005-11-28 | 2005-11-28 | Cigarette internal quality index extimating method based on regression function estimating SVM |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100444153C (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102823944A (en) * | 2012-05-24 | 2012-12-19 | 广东中烟工业有限责任公司 | Cigarette design adjusting method based on carbon monoxide emission load under deep inspiration condition |
Families Citing this family (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101419207B (en) * | 2008-10-27 | 2012-09-05 | 川渝中烟工业有限责任公司 | Method for predicting main index of flue-cured tobacco flume |
| CN102982383A (en) * | 2012-05-15 | 2013-03-20 | 红云红河烟草(集团)有限责任公司 | Energy supply and demand prediction method based on support vector machine |
| CN102823943B (en) * | 2012-05-24 | 2014-12-10 | 广东中烟工业有限责任公司 | Cigarette design adjusting method based on NNK release amount in deep smoking case |
| CN102697175B (en) * | 2012-05-24 | 2014-03-26 | 广东中烟工业有限责任公司 | Cigarette design adjusting method based on phenol discharging amount under deep-smoking condition |
| CN103336090B (en) * | 2013-05-14 | 2015-10-28 | 湖南中烟工业有限责任公司 | A kind of support vector machines method predicting Sensory Quality of Cigarette index |
| CN104809230A (en) * | 2015-05-08 | 2015-07-29 | 东北大学 | Cigarette sensory quality evaluation method based on multi-classifier integration |
| CN105606075B (en) * | 2015-12-18 | 2017-11-28 | 中国科学院南京地理与湖泊研究所 | A kind of method of discrimination of the large-scale shallow water lake wawter bloom Microcystis aeruginosa vertical characteristics pattern based on local wind speed |
| CN105954451B (en) * | 2016-06-06 | 2017-12-26 | 广东中烟工业有限责任公司 | Cigarette type Quick method based on the panchromatic modal data of electronic nose |
| CN107392399A (en) * | 2017-08-30 | 2017-11-24 | 桂林电子科技大学 | A kind of SVM Sensory Quality of Cigarette Forecasting Methodologies based on improved adaptive GA-IAGA |
| CN108323797B (en) * | 2018-02-09 | 2020-11-24 | 杭州安脉盛智能技术有限公司 | GPR (general purpose) model-based cigarette weight control system starting position prediction method and system |
| CN109919502A (en) * | 2019-03-13 | 2019-06-21 | 安徽海螺集团有限责任公司 | A kind of method for evaluating quality for uncertain data stream |
| CN110222061B (en) * | 2019-06-13 | 2023-08-25 | 红云红河烟草(集团)有限责任公司 | Equipment parameter management system and method for cigarette production line |
| CN111723858B (en) * | 2020-06-16 | 2023-07-11 | 中冶宝钢技术服务有限公司 | SVM-based stockpile quality estimation method and device, electronic equipment and medium |
| CN112749906A (en) * | 2021-01-14 | 2021-05-04 | 云南中烟工业有限责任公司 | Sensory evaluation method for spectrum data of cigarette mainstream smoke |
| CN112881323B (en) * | 2021-01-14 | 2022-03-08 | 云南中烟工业有限责任公司 | Quality evaluation method of cigarette mainstream smoke |
| CN114242183A (en) * | 2021-12-20 | 2022-03-25 | 甘肃烟草工业有限责任公司 | Cigarette smoke H value prediction model construction method based on tobacco chemistry convention |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6437317B1 (en) * | 1999-05-12 | 2002-08-20 | Focke & Co. (Gmbh & Co.) | Method and apparatus for inspecting cigarette heads |
| CN1115112C (en) * | 2000-12-21 | 2003-07-23 | 青岛海洋大学 | Method for creating fuzzy-neural network expert system for evaluating sensing quality of cigarette |
| CN1524460A (en) * | 2003-02-25 | 2004-09-01 | 中国海洋大学 | Method for establishing mixed expert system of maintaining cigarette leaf group formulation |
-
2005
- 2005-11-28 CN CNB2005101052856A patent/CN100444153C/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6437317B1 (en) * | 1999-05-12 | 2002-08-20 | Focke & Co. (Gmbh & Co.) | Method and apparatus for inspecting cigarette heads |
| CN1115112C (en) * | 2000-12-21 | 2003-07-23 | 青岛海洋大学 | Method for creating fuzzy-neural network expert system for evaluating sensing quality of cigarette |
| CN1524460A (en) * | 2003-02-25 | 2004-09-01 | 中国海洋大学 | Method for establishing mixed expert system of maintaining cigarette leaf group formulation |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102823944A (en) * | 2012-05-24 | 2012-12-19 | 广东中烟工业有限责任公司 | Cigarette design adjusting method based on carbon monoxide emission load under deep inspiration condition |
| CN102823944B (en) * | 2012-05-24 | 2014-12-10 | 广东中烟工业有限责任公司 | Cigarette design adjusting method based on carbon monoxide emission load under deep inspiration condition |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1975705A (en) | 2007-06-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100444153C (en) | Cigarette internal quality index extimating method based on regression function estimating SVM | |
| CN112288191B (en) | Ocean buoy service life prediction method based on multi-class machine learning method | |
| McGreal et al. | Neural networks: the prediction of residential values | |
| CN106407464A (en) | KNN-based improved missing data filling algorithm | |
| Lu et al. | Quality level identification of West Lake Longjing green tea using electronic nose | |
| CN108921604B (en) | Advertisement click rate prediction method based on cost-sensitive classifier integration | |
| CN108647730B (en) | Data partitioning method and system based on historical behavior co-occurrence | |
| CN101419209A (en) | Cigarette sensing appraise and flume index immune neural net prediction method | |
| CN112687349A (en) | Construction method of model for reducing octane number loss | |
| Hayden et al. | Statistical methods to develop rating models | |
| Rodriguez et al. | Time dependence of the aroma pattern emitted by an encapsulated essence studied by means of electronic noses and chemometric analysis | |
| CN114926299A (en) | Prediction method for predicting vehicle accident risk based on big data analysis | |
| CN111861667A (en) | Vehicle recommendation method and device, electronic equipment and storage medium | |
| CN113903407A (en) | Component identification method, component identification device, electronic equipment and storage medium | |
| CN116432856A (en) | Pipeline dynamic early warning method and device based on CNN-GLSTM model | |
| CN116881640A (en) | Method and system for predicting core extraction degree and computer-readable storage medium | |
| CN115630332A (en) | Wheat flour quality characteristic prediction method | |
| CN112378942B (en) | White spirit grade classification and identification method based on nuclear magnetic resonance fingerprint | |
| CN112733903B (en) | SVM-RF-DT combination-based air quality monitoring and alarming method, system, device and medium | |
| CN114757495A (en) | Membership value quantitative evaluation method based on logistic regression | |
| Terzi et al. | Comparison of financial distress prediction models: Evidence from turkey | |
| CN113064944A (en) | Data processing method and device | |
| Feng et al. | Intelligent techniques for cigarette formula design | |
| CN111415276B (en) | House combined renting recommendation method based on random forest | |
| Kubus | Assessment of Predictor importance with the example of the real estate market |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20081217 Termination date: 20131128 |