CN115015131A - Infrared spectrum training set sample screening method - Google Patents
Infrared spectrum training set sample screening method Download PDFInfo
- Publication number
- CN115015131A CN115015131A CN202210591768.5A CN202210591768A CN115015131A CN 115015131 A CN115015131 A CN 115015131A CN 202210591768 A CN202210591768 A CN 202210591768A CN 115015131 A CN115015131 A CN 115015131A
- Authority
- CN
- China
- Prior art keywords
- value
- negative
- samples
- positive
- intersection
- 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
Images
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
- G01N21/27—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands using photo-electric detection ; circuits for computing concentration
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Theoretical Computer Science (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)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention relates to the technical field of infrared spectrum data processing, and provides a method for screening infrared spectrum training set samples in order to improve the prediction accuracy of a quantitative model, which comprises the following steps: 1. centering the calibrated spectrum sample data; 2. acquiring a projection value of each spectrum sample data on a principal component coordinate axis; 3. dividing the spectrum sample data into four sets according to the positivity and the negativity of the projection value and the calibration value; 4. respectively acquiring the intersection of the four sets; 5. judging the correlation characteristics between the projection values and the calibration values according to the number of the samples in the four intersections; 6. determining a normal sample set and an abnormal sample set according to the correlation characteristics; 7. calculating the Mahalanobis distance corresponding to each sample spectrum value in the abnormal sample set; 8. and setting a Mahalanobis distance threshold, removing the samples with the Mahalanobis distance exceeding the threshold, and taking the rest samples as training set samples. The prediction accuracy of the quantitative model can be improved by adopting the steps.
Description
Technical Field
The invention relates to the technical field of infrared spectrum data processing, in particular to a method for screening samples in an infrared spectrum training set.
Background
The near infrared spectrum technology is widely applied to the detection of material components, and has the characteristics of rapidness, no damage, high efficiency, lower cost and the like. The portable near-infrared spectrometer refers to a near-infrared analysis instrument which can be held in a hand or carried by hand. The near-infrared spectrometer is portable, easy to carry, low in power consumption, high in detection speed and low in cost, and can be widely applied to multiple fields of agriculture, medicine, geology, environment and the like. However, the portable near-infrared spectrometer is easily affected by a light source, a detector, a detection method, environmental conditions and the like, so that the acquired spectral data has poor stability and low precision, and the spectral prediction analysis capability of the portable near-infrared spectrometer is further affected. In the practical application process, the spectrum data acquired by the portable near infrared spectrum equipment is easy to be abnormal, and the portable near infrared spectrum analysis technology is easy to be influenced by the abnormal spectrum data so as to greatly reduce the prediction analysis capability of the portable near infrared spectrum analysis technology. Meanwhile, when a portable spectrometer is used for collecting spectrum samples, due to manual operation, sample states, instrument states and the like, a small amount of abnormal samples exist in the obtained spectrum sample set, and when the spectrum sample set containing the abnormal samples is directly used as a training set to establish a quantitative model, the performance and the prediction capability of the model are reduced.
Disclosure of Invention
In order to improve the prediction accuracy of the quantitative model, the application provides a method for screening infrared spectrum training set samples.
The technical scheme adopted by the invention for solving the problems is as follows:
the infrared spectrum training set sample screening method comprises the following steps:
step 1, performing centralized processing on calibrated spectrum sample data;
step 2, acquiring a projection value of each spectrum sample data on a principal component coordinate axis;
step 3, dividing the spectrum sample data into a positive projection value set and a negative projection value set according to the positivity and negativity of the projection values; dividing the spectrum sample data into a positive calibration value set and a negative calibration value set according to the positive and negative of the calibration value after the centralization treatment;
and 4, respectively acquiring the intersection of the four sets: positive projection value positive calibration value intersection, positive projection value negative calibration value intersection, negative projection value positive calibration value intersection and negative projection value negative calibration value intersection;
step 5, judging the correlation characteristics between the projection value and the calibration value according to the number of the samples in the four intersections;
step 6, determining a normal sample set and an abnormal sample set according to the correlation characteristics;
step 7, calculating the Mahalanobis distance corresponding to each sample spectrum value in the abnormal sample set;
and 8, setting a Mahalanobis distance threshold, removing the samples of which the Mahalanobis distance exceeds the Mahalanobis distance threshold, and taking all the residual samples as training set samples.
Further, in the step 1, the centering method includes: and respectively subtracting the average spectrum data and the average calibration value of the training set samples from the spectrum data and the calibration value of each sample.
Further, in the step 5, if the number of samples of the intersection of the positive projection values and the positive calibration values + the number of samples of the intersection of the negative projection values and the negative calibration values is greater than the number of samples of the intersection of the positive projection values and the negative calibration values + the number of samples of the intersection of the negative projection values and the positive calibration values, the projection values of the samples and the calibration values are in positive correlation; and if the number of the samples of the positive projection value positive calibration value intersection and the number of the samples of the negative projection value negative calibration value intersection are less than the number of the samples of the positive projection value negative calibration value intersection and the number of the samples of the negative projection value positive calibration value intersection, the projection value of the samples and the calibration value are in negative correlation.
Further, in step 6, if the projection value of the sample and the calibration value are positively correlated, the normal sample set is: the positive projection value positive calibration value intersection and the negative calibration value intersection are respectively the following, and the abnormal sample set is as follows: the intersection of the positive projection value and the negative calibration value is a intersection of the positive calibration value and the negative calibration value; if the projection value of the sample and the calibration value are in negative correlation, the normal sample set is as follows: the intersection of the positive projection value and the negative calibration value and the intersection of the positive calibration value and the negative projection value are as follows, the abnormal sample set is as follows: the positive projection value positive calibration value intersection is the intersection of the negative calibration values of the negative projection values.
Further, in the step 8, the number of samples to be removed accounts for 20% of the abnormal sample set.
Compared with the prior art, the invention has the beneficial effects that: and determining abnormal sample points according to the correlation characteristics between the projection value and the calibration value of the sample data, and eliminating the abnormal sample points to ensure the accuracy of the spectral data in the training set, so that the accuracy of the model in training is improved, and the prediction accuracy of the quantitative model is further improved.
Drawings
FIG. 1 is a flow chart of a method for screening infrared spectroscopy training set samples;
FIG. 2 is a flow chart corresponding to the embodiment;
FIG. 3 is a schematic diagram of four subsets and their intersection.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
As shown in fig. 1, the method for screening a sample in an infrared spectrum training set comprises the following steps:
step 1, performing centralized processing on calibrated spectrum sample data;
step 2, acquiring a projection value of each spectrum sample data on a principal component coordinate axis;
step 3, dividing the spectrum sample data into a positive projection value set and a negative projection value set according to the positivity and negativity of the projection values; dividing the spectrum sample data into a positive calibration value set and a negative calibration value set according to the positive and negative of the calibration value after the centralization treatment;
and 4, respectively acquiring the intersection of the four sets: the positive projection value positive calibration value intersection, the positive projection value negative calibration value intersection, the negative projection value positive calibration value intersection and the negative projection value negative calibration value intersection;
step 5, judging the correlation characteristics between the projection value and the calibration value according to the number of the samples in the four intersections;
step 6, determining a normal sample set and an abnormal sample set according to the correlation characteristics;
step 7, calculating the Mahalanobis distance corresponding to each sample spectrum value in the abnormal sample set;
and 8, setting a Mahalanobis distance threshold, removing the samples of which the Mahalanobis distance exceeds the Mahalanobis distance threshold, and taking all the residual samples as training set samples.
Specifically, in the step 1, the centering method includes: and respectively subtracting the average spectral data and the average calibration value of the training set samples from the spectral data and the calibration value of each sample.
In the step 5, if the number of samples of the intersection of the positive projection values and the positive calibration values plus the number of samples of the intersection of the negative projection values and the negative calibration values is greater than the number of samples of the intersection of the positive projection values and the negative calibration values plus the number of samples of the intersection of the negative projection values and the positive calibration values, the projection values of the samples and the calibration values are in positive correlation; and if the number of the samples of the positive projection value positive calibration value intersection and the number of the samples of the negative projection value negative calibration value intersection are less than the number of the samples of the positive projection value negative calibration value intersection and the number of the samples of the negative projection value positive calibration value intersection, the projection value of the samples and the calibration value are in negative correlation. In step 6, if the projection value of the sample is positively correlated with the calibration value, the normal sample set is: the intersection of the positive projection values and the negative calibration values is U, and the abnormal sample set is as follows: the intersection of the positive projection value and the negative calibration value is a intersection of the positive calibration value and the negative calibration value; if the projection value of the sample and the calibration value are in negative correlation, the normal sample set is as follows: the intersection of the positive projection value and the negative calibration value is U, the intersection of the positive calibration value and the negative calibration value is U, and the abnormal sample set is as follows: the positive projection value positive calibration value intersection is the intersection of the negative calibration values of the negative projection values.
Preferably, in the step 8, the number of the rejected samples accounts for 20% of the abnormal sample set.
Examples
As shown in fig. 2, the spectral data of the training set is collected by a portable near-infrared spectrometer, the wavelength range of the spectral data is 1350nm to 2150nm, the resolution is 20nm, and each spectral data includes 300 wavelength points.
Centering the calibrated spectrum sample data to obtain centered spectrum data X0i and a calibrated value Y0 i: respectively subtracting the average spectral data value Xmean and the average calibration value Ymean of the training set samples from the spectral data Xi and the calibration value Yi of each sample, namely: x0i ═ Xi-Xmean; y0i ═ Yi-Ymean, i ═ 1,2, …, m, where m is the number of samples.
A quantitative model is constructed by utilizing a partial least square method, and in the process of establishing the model, each sample needs to be projected on the main component coordinate axes which are orthogonal with each other in sequence, so that the projection value Xsi of the first main component can be obtained.
The training set samples are divided into two subsets using the negativity or positivity of the first principal component in the projection values: a set of positive projection values P _ score _ set and a set of negative projection values N _ score _ set; the training set samples are also divided into two subsets by using the positivity and the negativity of the calibration values: a set of positive calibrations P _ Y _ set and a set of negative calibrations N _ Y _ set.
Four intersections of the four subsets are respectively obtained, as shown in fig. 3, a positive projection value positive calibration value intersection Ps _ Py _ set, a positive projection value negative calibration value intersection Ps _ Ny _ set, a negative projection value positive calibration value intersection Ns _ Py _ set, and a negative projection value negative calibration value intersection Ns _ Ny _ set are obtained, and the number of samples in the four subsets is num _ Ps _ Py _ set, num _ Ps _ Ny _ set, num _ Ns _ Ny _ set, and num _ Ns _ Py _ set.
Judging the correlation characteristic between the projection value and the calibration value according to the number of samples in the four intersections, wherein if the number of samples of the positive projection value and the positive calibration value intersection plus the number of samples of the negative projection value and the negative calibration value intersection is greater than the number of samples of the positive projection value and the negative calibration value intersection plus the number of samples of the negative projection value and the positive calibration value intersection, the projection value of the sample is positively correlated with the calibration value; and if the number of the samples of the positive projection value positive calibration value intersection and the number of the samples of the negative projection value negative calibration value intersection are less than the number of the samples of the positive projection value negative calibration value intersection and the number of the samples of the negative projection value positive calibration value intersection, the projection value of the samples and the calibration value are in negative correlation.
To obtain more accurate correlation, four correlation characteristics are adopted in the embodiment: the positive and negative relations between the projection value and the calibration value are positive correlation, negative correlation, weak positive correlation and weak negative correlation.
(1) When the number of samples in the positive projection value positive calibration value intersection is larger than that in the positive projection value negative calibration value intersection, and the number of samples in the negative projection value negative calibration value intersection is larger than that in the negative projection value positive calibration value intersection, namely (num _ Ps _ Py _ set > num _ Ps _ Ny _ set) & (num _ Ns _ Ny _ set > num _ Ns _ Py _ set), the positive-negative relation between the projection value of the samples and the calibration value is considered to be in positive correlation;
(2) when the number of samples in the positive projection value positive calibration value intersection is smaller than that in the positive projection value negative calibration value intersection, and the number of samples in the negative projection value negative calibration value intersection is smaller than that in the negative projection value positive calibration value intersection, namely (num _ Ps _ Py _ set < num _ Ps _ Ny _ set) & (num _ Ns _ Ny _ set < num _ Ns _ Py _ set), the positive-negative relation between the projection value of the samples and the calibration value is considered to be negative correlation;
and when the two characteristics are not met, the positive and negative relation between the projection value of the sample and the calibration value is not obvious, and the weak correlation relation between the projection value of the sample and the calibration value is judged.
(3) When the sum of the number of samples in the positive projected value positive calibration value intersection and the number of samples in the negative projected value negative calibration value intersection is larger than the sum of the number of samples in the positive projected value negative calibration value intersection and the number of samples in the negative projected value positive calibration value intersection, namely when (num _ Ps _ Py _ set + num _ Ns _ Ny _ set) > (num _ Ps _ Ny _ set + num _ Ns _ Py _ set), the positive-negative relation between the projected value of the samples and the calibration value is considered to be weak positive-negative relation;
(4) and when the sum of the number of samples in the positive projection value positive calibration value intersection and the number of samples in the negative projection value negative calibration value intersection is smaller than the sum of the number of samples in the positive projection value negative calibration value intersection and the number of samples in the negative projection value positive calibration value intersection, namely when (num _ Ps _ Py _ set + num _ Ns _ Ny _ set) < (num _ Ps _ Ny _ set + num _ Ns _ Py _ set), the positive-negative relation between the projection value of the samples and the calibration value is considered to be weak negative-correlation.
Dividing the training set samples into two subsets of a normal sample set (normal _ set) and an abnormal sample set (unormal _ set) according to four kinds of relevant characteristics of the projection values and the calibration values, wherein the division rule is as follows:
(1) when the positive and negative relations between the projection value and the calibration value of the sample are in positive correlation or weak positive correlation, the normal sample set is the union of two subsets, namely Ps _ Py _ set _ Ns _ Ny _ set, of the intersection of the positive projection value and the negative calibration value, and the number of the samples is (num _ Ps _ Py _ set + num _ Ns _ Ny _ set). The abnormal set is a union of two subsets of a positive projection value negative calibration value intersection and a negative projection value positive calibration value intersection, namely Ps _ Ny _ set U _ Ns _ Py _ set, and the number of samples is (num _ Ps _ Ny _ set + num _ Ns _ Py _ set);
(2) when the positive and negative relations between the projection value and the calibration value of the sample are negative or weakly negative, the normal set is the union of two subsets, namely Ps _ Ny _ set $ Ns _ Py _ set, of the intersection of the negative calibration value of the positive projection value and the positive calibration value of the negative projection value, and the number of the samples is (num _ Ps _ Ny _ set + num _ Ns _ Py _ set). The abnormal set is a union of two subsets of intersection of the positive calibration value and the negative calibration value of the projection value, namely Ps _ Py _ set ═ Ns _ Ny _ set, and the number of samples is (num _ Ps _ Py _ set + num _ Ns _ Ny _ set).
Calculating the Mahalanobis distance corresponding to each sample spectrum value in the abnormal sample set: averaging sample spectrum data in the abnormal sample set to obtain an average spectrum Xunnormal of the abnormal sample set mean Therefore, the mahalanobis distance between each sample spectrum data and the average spectrum can be calculated, and the calculation formula is as follows:
wherein X _ normal is j … (j is 1, …, P) is the jth sample in the abnormal sample set, P is the number of samples in the abnormal sample set, and S is X _ unnormal j The covariance matrix of (2) has a matrix size of P × P.
And setting a Mahalanobis distance threshold, screening out the sample points exceeding the threshold range as abnormal sample points, and taking the rest all samples as training set samples. Sample points exceeding the threshold range are used as edge samples, the contribution value of the edge samples to the model is higher than that of samples with small Mahalanobis distance, and the performance of the model can be effectively improved after the sample points are used as abnormal sample points to be screened out; the threshold is set according to actual conditions, and the number of abnormal points which are generally screened accounts for about 20% of the abnormal sample set.
Claims (5)
1. The infrared spectrum training set sample screening method is characterized by comprising the following steps:
step 1, performing centralized processing on calibrated spectrum sample data;
step 2, acquiring a projection value of each spectrum sample data on a principal component coordinate axis;
step 3, dividing the spectrum sample data into a positive projection value set and a negative projection value set according to the positivity and negativity of the projection values; dividing the spectrum sample data into a positive calibration value set and a negative calibration value set according to the positive and negative of the calibration value after the centralization treatment;
and 4, respectively acquiring the intersection of the four sets: the positive projection value positive calibration value intersection, the positive projection value negative calibration value intersection, the negative projection value positive calibration value intersection and the negative projection value negative calibration value intersection;
step 5, judging the correlation characteristics between the projection value and the calibration value according to the number of the samples in the four intersections;
step 6, determining a normal sample set and an abnormal sample set according to the correlation characteristics;
step 7, calculating the Mahalanobis distance corresponding to each sample spectrum value in the abnormal sample set;
and 8, setting a Mahalanobis distance threshold, removing the samples of which the Mahalanobis distance exceeds the Mahalanobis distance threshold, and taking all the residual samples as training set samples.
2. The method for screening the infrared spectrum training set sample according to claim 1, wherein in the step 1, the method for centralization comprises the following steps: and respectively subtracting the average spectral data and the average calibration value of the training set samples from the spectral data and the calibration value of each sample.
3. The method for screening samples in an infrared spectroscopy training set according to claim 1, wherein in the step 5, if the number of samples in the intersection of the positive projection values and the positive calibration values + the number of samples in the intersection of the negative projection values and the negative calibration values > the number of samples in the intersection of the positive projection values and the negative calibration values + the number of samples in the intersection of the negative projection values and the positive calibration values, the projection values of the samples and the calibration values are positively correlated; and if the number of the samples of the positive projection value positive calibration value intersection and the number of the samples of the negative projection value negative calibration value intersection are less than the number of the samples of the positive projection value negative calibration value intersection and the number of the samples of the negative projection value positive calibration value intersection, the projection value of the samples and the calibration value are in negative correlation.
4. The method for screening samples in an infrared spectroscopy training set according to claim 3, wherein in the step 6, if the projection value of the sample and the calibration value are positively correlated, the normal sample set is: the positive projection value positive calibration value intersection and the negative calibration value intersection are respectively the following, and the abnormal sample set is as follows: the intersection of the positive projection value and the negative calibration value is a intersection of the positive calibration value and the negative calibration value; if the projection value of the sample and the calibration value are in negative correlation, the normal sample set is as follows: the intersection of the positive projection value and the negative calibration value and the intersection of the positive calibration value and the negative projection value are as follows, the abnormal sample set is as follows: the positive projection value positive calibration value intersection is the intersection of the negative calibration values of the negative projection values.
5. The method for screening a training set of infrared spectra as claimed in any of claims 1 to 4 wherein in step 8 the number of samples removed is 20% of the abnormal set of samples.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210591768.5A CN115015131A (en) | 2022-05-27 | 2022-05-27 | Infrared spectrum training set sample screening method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210591768.5A CN115015131A (en) | 2022-05-27 | 2022-05-27 | Infrared spectrum training set sample screening method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115015131A true CN115015131A (en) | 2022-09-06 |
Family
ID=83071424
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210591768.5A Pending CN115015131A (en) | 2022-05-27 | 2022-05-27 | Infrared spectrum training set sample screening method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115015131A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116818687A (en) * | 2023-06-21 | 2023-09-29 | 浙江大学杭州国际科创中心 | Soil organic carbon spectrum prediction method and device based on spectrum guide integrated learning |
-
2022
- 2022-05-27 CN CN202210591768.5A patent/CN115015131A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116818687A (en) * | 2023-06-21 | 2023-09-29 | 浙江大学杭州国际科创中心 | Soil organic carbon spectrum prediction method and device based on spectrum guide integrated learning |
| CN116818687B (en) * | 2023-06-21 | 2024-02-20 | 浙江大学杭州国际科创中心 | Soil organic carbon spectrum prediction method and device based on spectrum guide integrated learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106815643B (en) | Infrared spectroscopy Model Transfer method based on random forest transfer learning | |
| CN109799195B (en) | High-precision quantitative analysis method for laser-induced breakdown spectroscopy | |
| CN110717368A (en) | Qualitative classification method for textiles | |
| CN114168906B (en) | Mapping geographic information data acquisition system based on cloud computing | |
| CN103175805B (en) | Method for determining indexes of COD and BOD5 in sewage through near infrared spectrometry | |
| CN112417370B (en) | Mu Leqiong S matrix estimation and polarization noise analysis method for rough surface substances | |
| WO2018184262A1 (en) | Dynamic calibration method for echelle spectrometer for laser induced breakdown spectrum collection | |
| McNeish et al. | The effect of measurement quality on targeted structural model fit indices: A comment on Lance, Beck, Fan, and Carter (2016). | |
| CN108398416A (en) | A kind of mix ingredients assay method based on laser Raman spectroscopy | |
| CN115015131A (en) | Infrared spectrum training set sample screening method | |
| CN114611582B (en) | Method and system for analyzing substance concentration based on near infrared spectrum technology | |
| US20220042962A1 (en) | Method for validation of site-specific water quality criteria of river basin | |
| CN116187861A (en) | Isotope-based water quality traceability monitoring method and related device | |
| Sun et al. | A classification and location of surface defects method in hot rolled steel strips based on YOLOV7 | |
| CN113484400B (en) | Mass spectrogram molecular formula calculation method based on machine learning | |
| CN115392375A (en) | Intelligent evaluation method and system for multi-source data fusion degree | |
| CN114878544A (en) | Method for identifying target component from mixture SERS spectrum | |
| CN111426648B (en) | Method and system for determining similarity of infrared spectrogram | |
| CN111044504B (en) | Coal quality analysis method considering uncertainty of laser-induced breakdown spectroscopy | |
| CN109359678B (en) | High-precision classification recognition algorithm for liquor atlas | |
| CN113588597A (en) | Method for improving analysis precision of furnace slag | |
| CN110458228A (en) | A kind of hazardous material detection method of information source number time-varying and self-adaptive blind source separation | |
| CN116166973A (en) | Portable near infrared spectrum quantitative model and unknown sample matching degree judging method | |
| CN115931828B (en) | Component analysis and prediction method, unit and system suitable for complex soil matrix | |
| CN117420080B (en) | Water quality analysis method integrating planktonic algae pigment protein spectrum and image |
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 |