Methods for determining a binary classifier and methods for assigning a sample to one of two possible classes based on spectroscopic data of the sample.

Abstract

The invention relates to a computer-implemented method for determining a binary classifier, wherein the classifier is configured to assign a sample to one of two possible classes (P1, P2) based on spectroscopic data of a sample, comprising the steps: - Receiving two-dimensional spectroscopic data of the sample, wherein the data comprise discrete values ​​in a wavelength dimension (12) and a time dimension (14), - Determining features F1 to F5 formed along the time dimension (14) of the data for several wavelengths and / or for several wavelength ranges (16), - Determining linear combinations of features F1 to F5 for the multiple wavelengths and / or for the multiple wavelength ranges (16) by analyzing the features with respect to their power of separation for the two classes (P1, P2), and - Determining a subset of the identified linear combinations by analyzing the linear combinations with regard to their discriminatory power for the two classes (P1, P2). The invention further relates to a computer-implemented method for assigning a sample to one of two possible classes based on spectroscopic data of the sample. Furthermore, the invention relates to a device for data processing, a computer program product, and a computer-readable data carrier.

Description

[0001] The invention relates to a computer-implemented method for determining a binary classifier.

[0002] The invention further relates to a computer-implemented method for assigning a sample to one of two possible classes based on spectroscopic data of the sample.

[0003] Furthermore, the invention relates to a device for data processing, a computer program product, and a computer-readable data carrier.

[0004] Currently, there is an effort to determine the sex of a future chick as soon as it hatches from the fertilized chicken egg.

[0005] Different fluorophores develop in male and female chicks during the development of a bird's egg. Due to their complex structure, these molecules exhibit unpredictable fluorescence. Fluorescence is the observed energy transition from the excited state to the ground state of the molecule. This process is time-dependent.

[0006] Document WO 2021 / 144420A1 describes a device and a method for optical in-ovo sex determination in a fertilized bird egg. The device comprises a light source for emitting excitation radiation to excite fluorescence in a region inside the bird egg, a spectroscopic device for time- and / or spectrally resolved analysis of fluorescence radiation emitted from the region inside the bird egg, and an evaluation unit for sex determination from the data obtained by means of the spectroscopic device.

[0007] Document WO 2023 / 161532A1 relates to a method and apparatus for detecting and / or classifying particles of organic compounds from a fingerprint of backscattered light in a liquid dispersion sample. The method uses an electronic data processor for detecting and / or classifying particles in the sample, and includes using the electronic data processor to pretrain a machine learning classifier with a large number of sample particles.

[0008] The publication by Nipuna Rajapaksha et al., "Supervised machine learning algorithm selection for condition monitoring of induction motors," from Proceedings: 2021 IEEE Southern Power Electronics Conference (SPEC), December 6-9, 2021 - ISBN 978-1-6654-3623-6, describes supervised machine learning (ML) algorithms that can be used for condition monitoring of three-phase induction motors, as well as their advantages and disadvantages. It also discusses how the dominant features can be selected from raw data using time-domain and frequency-domain analysis of acoustic data collected from a three-phase induction motor. Furthermore, the document describes the classification accuracy of each ML algorithm and a procedure for selecting an algorithm based on experimental results.

[0009] Based on this, the object of the invention is to provide means to increase the accuracy of in-ovo sex determination and / or to improve the robustness of the method.

[0010] The problem is solved according to the invention by the features of the independent claims. Preferred embodiments of the invention are specified in the dependent claims, each of which, individually or in combination, can represent an aspect of the invention.

[0011] According to the invention, a computer-implemented method for determining a binary classifier is provided, wherein the classifier is configured to assign a sample to one of two possible classes based on spectroscopic data of a sample, comprising the steps - Receiving two-dimensional spectroscopic data of the sample, wherein the data comprise discrete values ​​in a wavelength dimension and a time dimension, - Determining features F1 to F5 formed along the time dimension of the data for multiple wavelengths and / or for multiple wavelength ranges, wherein the features • F1: Statistical moments of the 2nd to 4th order of the data in the time dimension, • F2: Coefficients of a regression line of the data in the time dimension, • F3: Real coefficients of a Fourier transform of the data in the time dimension, • F4: Coefficients of a regression line through the real coefficients of the Fourier transform determined for feature F3, and • F5: Include the entropy of the data in the time dimension, - Determining linear combinations of features F1 to F5 for the multiple wavelengths and / or for the multiple wavelength ranges by analyzing the features with regard to their discriminatory power for the two classes, and - Determining a subset of the identified linear combinations by analyzing the linear combinations with regard to their discriminatory power for the two classes.

[0012] Furthermore, the invention relates to a computer-implemented method for assigning a sample to one of two possible classes based on spectroscopic data of the sample, and in particular for assigning a fertilized bird egg to one of two possible sexes, comprising the steps - Receiving two-dimensional spectroscopic data of the sample, wherein the data comprise discrete values ​​in a wavelength dimension and a time dimension, and - Applying a classifier to the received data, where the classifier was determined using the above methods.

[0013] Furthermore, the invention relates to a device for data processing comprising means for carrying out one of the two above methods or means for carrying out both methods.

[0014] Furthermore, the invention relates to a computer program product comprising instructions which, when the program is executed by a computer, cause it to execute one or both of the above methods.

[0015] Furthermore, the invention relates to a computer-readable data carrier on which the above computer program product is stored.

[0016] One aspect of the invention is that it has been found that when assigning a sample to one of two possible classes based on spectroscopic data of the sample, an improved hit rate is achieved if not only statistical features of the spectroscopic data are considered to determine the classifier in the training phase.

[0017] Another aspect of the invention is the use of a two-stage method for determining the classifier. It has been found that a classifier that achieves a higher hit rate in sample assignment is obtained by determining the linear combinations of features F1 to F5 for the multiple wavelengths and / or for the multiple wavelength ranges by analyzing the features with respect to their discriminatory power for the two classes, and by determining a subset of the determined linear combinations by analyzing these combinations with respect to their discriminatory power for the two classes. In other words, a model is built using the training data in a two-stage process.

[0018] Furthermore, a classifier determined according to the inventive method can provide a significantly more robust classification method, in which samples whose spectroscopic data were not measured under ideal conditions and exhibit an increased signal-to-noise ratio can nevertheless be assigned to the two possible classes with a high success rate. Preferably, the classifier is a linear classifier. A linear classifier separates the classes along a linear hyperplane.

[0019] The two-dimensional spectroscopic data of the sample are preferably stored in a two-dimensional data matrix, wherein one row of the data matrix preferably comprises the time response, and in particular the decay curves, for individual singular wavelengths, or an averaged time response, and in particular averaged decay curves, for several directly adjacent singular wavelengths. In the second case, one row of the data matrix thus corresponds to the time response of a wavelength range. The wavelength range preferably has a bandwidth of no more than 4 nm, preferably no more than 3 nm. It is further preferably provided that the number of adjacent singular wavelengths averaged for the wavelength range is no more than eight, and preferably no more than six.

[0020] In the first stage of the procedure for determining the binary classifier, the features F1 to F5 are first determined for the rows of the matrix. Preferably, the features F1 to F5 are determined for each row of the matrix. In other words, the features F1 to F5 of singular wavelengths and / or of wavelength ranges are preferably formed.

[0021] With regard to feature F1, the statistical moments of the 2nd to 4th order are preferably the central moments, i.e. the central moment of the 2nd order (standard deviation), the central moment of the 3rd order (skewness) and the central moment of the 4th order (kurtosis).

[0022] With respect to feature F2, the coefficients of the regression line are preferably the coefficients of a normal equation g(x) = ax + b through the data in the time dimension.

[0023] With respect to feature F3, the real coefficients are preferably the real coefficients of the Fast Fourier Transform of the data in the time dimension.

[0024] For feature F4, the coefficients of the regression line are preferably the coefficients of a normal equation f(x) = sx + t through the real coefficients of the Fast Fourier transform determined for feature F3.

[0025] With respect to feature F5, the entropy of the data in the time dimension is preferably the Shannon entropy according to E = - Σ z p z log2 p z , with p z as probability for the measured value e.g.

[0026] The features are then analyzed with regard to their resolution, in order to preferentially eliminate features with weak resolution. In other words, linear combinations are determined from the features for each wavelength and / or wavelength range, whereby the eliminated features have a coefficient of zero in the linear combination and are therefore not considered. After this step, there is thus preferably a linear combination of features F1 to F5 specific to each row of the data matrix.

[0027] In the subsequent step of the process, a subset of the linear combinations generated is determined in the second stage. This is done by analyzing the linear combinations with regard to their separation power in order to preferentially eliminate linear combinations with weak separation power. In other words, certain linear combinations with particularly high separation powers are preferentially identified.

[0028] In other words, it is preferred that the classifier be determined by feature selection and / or selection of linear combinations based on training data. In particular, it has been shown that the two-stage method can yield a binary classifier with a particularly high hit rate. The classifier is also particularly robust, as it enables the reliable classification of spectroscopic data from the sample with a high signal-to-noise ratio.

[0029] According to a preferred embodiment of the invention, the sample is a fertilized bird egg, and the two classes represent a male and a female sex of the fertilized bird egg. The method for determining the classifier has proven particularly suitable for determining a classifier for determining the sex of the fertilized bird egg. In particular, with the classifier determined according to the present method, hit rates of 100% could be achieved in the classification of the fertilized bird egg.

[0030] According to a further preferred embodiment of the invention, the received two-dimensional spectroscopic data are intrinsic fluorescence data of the sample, and in particular intrinsic fluorescence data of the sample, especially of the fertilized bird egg, acquired by means of time-resolved laser-induced fluorescence spectroscopy (zLIF) and / or time-correlated single-photon counting (TCSPC). In other words, with regard to determining a binary classifier, which is preferably used for in-ovo sex determination in a fertilized bird egg, knowledge of the lifetime and decay profile of excited molecular states (time dimension of the two-dimensional data) in addition to the energy of the emitted photons (wavelength dimension of the two-dimensional data) is relevant for identifying the sex of the fertilized bird egg.

[0031] To detect intrinsic fluorescence, spectroscopic data can be acquired using laser-induced fluorescence techniques. This method relies on the fluorescence excitation of the sample by an excitation pulse from a light source such as a laser or an LED. The fluorophores present in the sample, and preferably in the bird's egg, are excited by the laser. After a certain time, typically on the order of a few nanoseconds to microseconds, the excited fluorophores will decay and emit light with a wavelength longer than the excitation wavelength. This fluorescence light is typically recorded using a photomultiplier tube (PMT) or a multichannel detector configured as an ICCD camera.

[0032] In some methods - also known as boxcar methods - the complete spectrum is recorded at different times after an excitation pulse using an ICCD camera.

[0033] Time-correlated single-photon counting (TCSPC) can also be used to detect autofluorescence. With TCSPC, the entire spectrum is not recorded after each excitation pulse. Instead, individual photons of a periodic light signal—in this case, the autofluorescence—are detected, and the respective times between the excitation pulse of the pulsed excitation radiation and the arrival of the photon at the detection device are determined. In other words, the time measurement is started by the excitation pulse, and the photon emitted during the transition from the excited state to the ground state stops the measurement. The measurement is repeated many times, and the individual time-correlated photons (with respect to the excitation pulse) are sorted into a so-called TCSPC histogram according to their measured time. The TCSPC histogram represents the temporal evolution of the autofluorescence radiation after excitation.Preferably, the TCSPC histogram generated by the detection device has a bin width, also called class width, for histogram classes from 1 ps to 50 ps, ​​preferably from 10 ps to 20 ps. Preferably, the bin width of the TCSPC histogram can be adapted to the device and / or the sample under investigation. Furthermore, preferably, when adapting the bin width of the TCSPC histogram, a temporal resolution of the entire device—and particularly preferably, a full width at half maximum (FWHM) of the instrument response function (IRF)—is taken into account. The FWHM of the IRF depends essentially on the light source and a pulse length generated by the light source and / or on a detector element of the detection device.

[0034] According to a preferred embodiment of the invention, the data for determining features F2 to F5 are provided as normalized data in the time dimension, such that the mean of the data in the time dimension is zero. It is further preferred that, for determining one or more of the features F2 to F5, the data are also normalized such that the standard deviation of the data in the time dimension is one. It has been shown that better results for classification are achieved if the raw data provided by the measurement are not used directly to determine features F2 to F5. Instead, it is advantageous if the time dimension of the data is provided in normalized form, where the mean µ = 0 and, preferably, the standard deviation σ = 1 for some or all features F2 to F5.

[0035] In contrast, according to a further preferred embodiment of the invention, it is provided that the data for determining feature F1 are available as unnormalized data in the time dimension. In other words, the unnormalized raw data are preferably used to determine the standard deviation, skewness, and kurtosis.

[0036] In connection with features F3 and F4, a further embodiment of the invention provides that the real coefficients of the Fourier transform of the data in the time dimension of feature F3 are ordered in ascending order, and / or that the coefficients of the regression line are determined by the real coefficients of the Fourier transform determined for feature F3, based on ascendingly ordered real coefficients of the Fourier transform. In other words, for feature F3, the real coefficients of the Fourier transform are sorted—namely, in ascending order. Furthermore, in connection with feature F4, the regression line f(x) = sx + t is preferably formed by the ascendingly ordered real coefficients of the Fourier transform.

[0037] In connection with the determination of linear combinations, a further development of the method provides that the determination of linear combinations of features F1 to F5 includes determining the coefficients of the linear combinations. The linear combinations are preferably linear combinations according to ∑i=1i=5ciFi, where c i the respective coefficient of the characteristic F i is. The coefficients c iThese values ​​can be arbitrary. If, through analysis of the features with regard to their resolution, it is determined that a feature does not have a high resolution, it is preferably provided that the corresponding coefficient of the feature in the linear combination is 0, and the feature is eliminated accordingly. In this context, it is preferably provided that the step of determining linear combinations of features F1 to F5 for the multiple wavelengths and / or for the multiple wavelength ranges includes defining a coefficient of the linear combination as zero for features whose resolution is below a predefined threshold.

[0038] According to a further preferred embodiment of the invention, the step of determining the linear combinations of features F1 to F5 for the multiple wavelengths and / or for the multiple wavelength ranges by analyzing the features with respect to their discriminatory power for the two classes; and / or the step of determining the subset of the determined linear combinations by analyzing the linear combinations with respect to their discriminatory power for the two classes comprises analysis by means of linear discriminant analysis. In other words, the measured spectroscopic data are preferably processed by feature engineering to determine the classifier, whereby features and linear combinations formed from the features with weak discriminatory power are eliminated.It has been shown that a classifier determined without eliminating features and / or linear combinations has a lower hit rate than one determined with feature elimination and / or linear combination elimination.

[0039] In this context, according to a further preferred development, it is provided that a discriminant function, and preferably the Fisher discriminant function and / or a hit rate of the determined classifier, is used as a metric for the separation strength during the analysis.

[0040] As already mentioned, the invention also relates to a method for assigning a sample to one of two possible classes based on spectroscopic data of the sample, and in particular for assigning a fertilized bird egg to one of two possible sexes, comprising the steps - Receiving two-dimensional spectroscopic data of the sample, wherein the data comprise discrete values ​​in a wavelength dimension and a time dimension, and - Applying a classifier to the received data, the classifier being determined using the method described above.

[0041] Preferably, the sample is a fertilized bird egg. In this method, the sex determination of the bird egg is preferably carried out by prioritizing specific linear combinations, wherein the linear combinations were formed from the characteristics F1 to F5 derived from the time dimension of the data.

[0042] The person skilled in the art will derive further technical aspects and advantages of the method for assigning a sample to one of two possible classes based on spectroscopic data of the sample from the above description of the method for determining the binary classifier.

[0043] The invention is explained below by way of example with reference to the accompanying drawings and preferred embodiments, wherein the features shown below can represent an aspect of the invention, either individually or in combination. The drawings show: Fig.1 in a) a schematic representation of two-dimensional spectroscopic data of a sample obtained in a method for determining a binary classifier according to a preferred embodiment of the invention, and in b) a schematic representation of a separation strength of features formed along the time dimension of the data in feature space, and Fig. 2 a schematic representation of a linear discriminant function which, within the framework of a method for determining a binary classifier according to a preferred embodiment of the invention, separates linear combinations of the features formed along the time dimension of the data into two classes.

[0044] Fig. 1 shows left in Fig.1a) a schematic representation of two-dimensional spectroscopic data of a sample obtained in a method for determining a binary classifier according to a preferred embodiment of the invention.

[0045] The spectroscopic data presented here are time-resolved fluorescence emissions measured on hatching eggs on the third day of incubation. The sex of the hatching eggs was determined by PCR.

[0046] Time-resolved fluorescence emissions were recorded using the Boxcar measurement method with a fluorescence excitation of 266 nm and the following parameters: Emission parameters: τ min = 0 ns; τ max = 30 ns (Discrediting = 1 ns); λ min = 340.14 nm; λ max = 720.60 nm (discreditation 0.5 nm).

[0047] The two-dimensional spectroscopic data of the sample received for determining the classifier are presented here in a two-dimensional data matrix A, where one row (a j1 , ..., a jn ), j = 1, ... ,m of the data matrix includes the time behavior and in particular the decay curves 10a for individual singular wavelengths or an averaged decay curve 10b for several directly adjacent singular wavelengths. A=(a11a12⋯a1na21a22…a2n⋮⋮⋯⋮am1am2⋯amn),

[0048] When the data matrix A is represented graphically, the result is as follows: Fig. 1 a) The image shown, where the y-axis 12 corresponds to the wavelength dimension and the x-axis 14 to the time dimension, shows the decay curves 10a of individual singular wavelengths as well as decay curves 10b averaged over the wavelength range 16.

[0049] In the preferred embodiment of the method for determining the binary classifier described below, features F1 to F5 formed along the time dimension of the data are determined for several wavelengths and / or for several wavelength ranges after the two-dimensional data have been received.

[0050] The characteristics F1 to F5 are as follows: • F1: Statistical moments of the 2nd to 4th order of the data in the time dimension, • F2: Coefficients of a regression line of the data in the time dimension, • F3: Real coefficients of a Fourier transform of the data in the time dimension, • F4: Coefficients of a regression line through the real coefficients of the Fourier transform determined for feature F3, and • F5: Entropy of the data in the time dimension

[0051] Since the decay curves 10 are a function of time, they will also be referred to as S(τ) in the following. The decay curves 10 are received in an original, unnormalized form.

[0052] The feature F1 is therefore the standard deviation σ, skewness γ and kurtosis ω calculated for S(τ).

[0053] To determine the features F2 to F5, the decay curves are converted into a normalized form S N (τ) brought, where S(τ) has the mean µ = 0 and the standard deviation σ = 1.

[0054] Feature F2 refers to the coefficients a, b of the normal equation g(x) = ax + b for S N (τ).

[0055] Feature F3 consists of the ascendingly ordered real coefficients of the Fast Fourier Transform FFT[S]. N (τ)].

[0056] Feature F4 refers to the coefficients s, t of the normal equation f(x) = sx + t for the ascendingly ordered real coefficients of the Fast Fourier Transform FFT[S N (τ)].

[0057] Feature F5 is the entropy E [S N (τ)].

[0058] Subsequently, linear combinations of the features F1 to F5, which are also referred to as profiles pr in the following, are calculated according to pr=∑i=1i=5ciFi formed, where c i the respective coefficient of the characteristic F i is.

[0059] To determine the profiles, the characteristics are analyzed with regard to their distinguishing strength for the two classes. Fig. Figure 1b) shows, by way of example, the result of the analysis in feature space for the decay curves 10a for singular wavelengths and for the averaged decay curve 10b.

[0060] In the present embodiment, features exhibiting a separation strength below a predefined threshold are represented in the linear combination as the value for the coefficient c. i = 0 assigned, so that in other words they are eliminated, while the other characteristics of the coefficients have a value c. i = 1 is assigned.

[0061] This procedure is illustrated here for the profile around 524.38 nm: For this purpose, individual spectra between 518.80 nm and 529.95 nm were analyzed using machine learning methods for the profile around 524.38 nm. It turns out that features F1, F2, and F4 do not have sufficient discrimination for this profile. Further consideration of these features would worsen the classification accuracy. The following values ​​for features F3 and F5, which must be considered in the linear combination, are obtained: F3: - Four Fourier transform coefficients: f1, f2, f3, f4 at 523.82 nm - Three Fourier transform coefficients: f5, f6, f7 at 523.26 nm - A Fourier transform coefficient: f8 at 524.38 nm F5: - Entropy E1 at 518.80 nm - Entropy E2 at 527.17 nm - Entropy E3 at 529.95 nm - Entropy E4 at 523.26 nm

[0062] The profile pr=∑i=1i=5ciFi At a wavelength of 524.38 nm, the following coefficients c are therefore calculated i c1=0, c2=0, c3=1, c4=0 c5=1 according to pr524.38nm=c3(20f1−13f2−6f3−4.6f4+7.8f5−6.3f6+1.7f7+2.5f8) +c5(−8E1+7E2−4.5E3+4.8E4)−3.5

[0063] The value -3.5 also specifies that a classification rule states that >0 applies to one class and <0 to the other class.

[0064] Following this first stage of the procedure, the classification results listed below can be achieved for the following exemplary profiles. The hit rates were verified using two-dimensional spectroscopic data from hatching eggs that were not used in the classifier's training program. Profile at wavelength / nm Discriminant D Hit rate A / % 488,1183 0,43371 72,4567 490,9085 0,39672 72,0779 493,6985 0,53655 72,4567 496,4885 0,97459 75,7576 499,2783 0,86616 79,4371 502,0681 0,77513 75,7576 504,8577 0,68069 75,7576 507,6473 0,6117 72,4567 510,4367 0,9705 79,0584 513,2261 0,53488 72,4567 516,0153 1,1332 79,4372 518,8044 0,83231 79,4372 521,5934 0,73678 75,7576 524,3823 2,3048 90,0974 527,1710 0,89419 79,4372 529,9597 0,53765 72,4567 532,7482 0,9723 79,4372 535,5366 1,1829 79,4372 538,3249 0,56901 72,4567 541,1130 0,63482 75,3788 543,9011 1,2399 82,7381 546,6890 0,7954 75,0000 549,4768 1,0302 75,75761 552,2644 0,90166 78,6797 555,0519 0,82719 75,3788 557,8393 0,94714 79,0584 ...

[0065] D denotes the value of the discriminant function for the profiles, and A denotes the accuracy of the profiles. The wavelength designation corresponds either to the single, singular wavelength of the profile or to the arithmetic mean of the initial and final values ​​of the wavelength range.

[0066] In the next step of the procedure, a subset is selected from the determined profiles by analyzing the profiles with regard to their discrimination power for the two classes. This is done using linear discriminant analysis. Fig.Figure 2 shows schematically how the profiles represented as circles 18 are separated into two classes P1 and P2 by a discriminant function 20 found by means of linear discriminant analysis.

[0067] Taking into account this identified subset of profiles, which also includes the profile at 524.38 nm, a 100% hit rate is achieved in the 2nd stage. Reference sign 10a Decay curve for a single wavelength 10b Averaged decay curve over wavelength range 12 y-axis 14 x-axis 16 wavelength range 18 Circle 20 Discriminant function P1, P2 classes

Claims

[1] Computer-implemented method for determining a binary classifier, wherein the classifier is designed to assign a sample to one of two possible classes (P1, P2) on the basis of spectroscopic data of this sample, comprising the steps: - receiving two-dimensional spectroscopic data of the sample, the data comprising discrete values in a wavelength dimension (12) and a time dimension (14), - Determining features F1 to F5 formed along the time dimension (14) of the data for a plurality of wavelengths and / or for a plurality of wavelength ranges (16), wherein the features • F1: statistical moments of the 2nd to 4th order of the data in the time dimension (14), • F2: coefficients of a best fit line of the data in the time dimension (14), • F3: Real coefficients of a Fourier transform of the data in the time dimension (14), • F4: coefficients of a best fit line through the real coefficients of the Fourier transform determined for feature F3, and • F5: Entropy of the data in the time dimension (14) include, - determining linear combinations of the features F1 to F5 for the plurality of wavelengths and / or for the plurality of wavelength ranges (16) by analyzing the features with regard to their separation strength for the two classes (P1, P2), and - Determine a subset of the determined linear combinations by analyzing the linear combinations with regard to their discriminatory power for the two classes (P1, P2). [2] The method according to claim 1, wherein the sample is a fertilized bird egg and the two classes (P1, P2) represent a male sex and a female sex of the fertilized bird egg. [3] Method according to one of the preceding claims, wherein the received two-dimensional spectroscopic data are intrinsic fluorescence data of the sample and in particular intrinsic fluorescence data of the sample acquired by means of time-resolved laser-induced fluorescence spectroscopy (zLIF) and / or by means of time-correlated single photon counting (TCSPC). [4] Method according to one of the preceding claims, wherein for determining the features F2 to F5 the data in the time dimension (14) are present as normalized data such that a mean value of the data in the time dimension (14) is zero. [5] Method according to one of the preceding claims, wherein for determining the feature F1 the data in the time dimension (14) are present as non-normalized data. [6] Method according to one of the preceding claims, wherein the real coefficients of the Fourier transform of the data in the time dimension (14) of the feature F3 are ordered in ascending order and / or wherein the coefficients of the best-fit line are determined by the real coefficients of the Fourier transform determined for feature F3 on the basis of ascending order real coefficients. [7] Method according to one of the preceding claims, wherein the step of determining the linear combinations of the features F1 to F5 for the plurality of wavelengths and / or for the plurality of wavelength ranges (16) by analyzing the features with regard to their separation strength for the two classes (P1, P2) comprises defining a coefficient of the linear combination as zero for features whose separation strength is below a predefined threshold. [8] Method according to one of the preceding claims, wherein the step of determining the linear combinations of the features F1 to F5 for the plurality of wavelengths and / or for the plurality of wavelength ranges (16) by analyzing the features with regard to their separation strength for the two classes (P1, P2), and / or wherein the step of determining the subset of the determined linear combinations by analyzing the linear combinations with regard to their separation strength for the two classes (P1, P2) comprises analyzing by means of linear discriminant analysis. [9] Method according to the preceding claim, wherein a discriminant function (20) and preferably the Fisher discriminant function and / or a hit rate of the determined classifier are used as a metric for the separation strength during the analysis. [10] A computer-implemented method for classifying a sample into one of two possible classes (P1, P2) on the basis of spectroscopic data of the sample, and in particular for classifying a fertilized bird egg into one of two possible sexes, comprising the steps of - receiving two-dimensional spectroscopic data of the sample, the data comprising discrete values in a wavelength dimension (12) and a time dimension (14), and - applying a classifier to the received data, wherein the classifier was determined using the method according to one of claims 1 to 8. [11] Device for data processing comprising means for carrying out the method according to one of the preceding method claims. [12] Computer program product comprising instructions which, when the program is executed by a computer, cause the computer to carry out the method according to one of the preceding method claims. [13] A computer-readable data carrier on which the computer program product according to the preceding claim is stored.

Images