Assessment method and device for bee breed conservation effect, electronic equipment and storage medium

Abstract

The invention provides an evaluation method and device for a bee breed conservation effect, electronic equipment and a storage medium. The method comprises the steps of obtaining a plurality of original morphological indexes of to-be-detected bees of a target variety; inputting the plurality of original morphological indexes of the to-be-detected bee variety into a pre-trained XGBoost model for learning analysis, and outputting a plurality of target morphological indexes corresponding to the target variety by the XGBoost model; and evaluating the variety of the to-be-detected bees according to the target morphological indexes to obtain an evaluation result of the breed conservation effect of the to-be-detected bees. By means of the XGBoost model, intelligent screening of key morphological indexes and standardization of an evaluation process are achieved, rapid detection is achieved, and the efficiency and scientificity of breed conservation work are greatly improved.

Description

Technical Field

[0001] This application relates to the field of beekeeping technology, and more specifically, to a method, apparatus, electronic device, and storage medium for evaluating the effectiveness of bee conservation. Background Technology

[0002] Honeybees, as one of the world's most important pollinating insects, play an irreplaceable role in maintaining ecosystem balance and ensuring agricultural production. Among them, the Western honeybee, due to its efficient pollination ability and good adaptability, is widely used in modern agriculture; while native bee species, such as the Hunchun black bee, possess excellent cold resistance, disease resistance, and a high degree of adaptability to the local environment, making them valuable germplasm resources. With changes in the ecological environment, the introduction of alien bee species, and the intensive development of beekeeping activities, native bee species face serious threats such as genetic erosion, hybridization pollution, and population decline. Therefore, carrying out scientific and effective bee conservation work has become an urgent need for biodiversity protection and sustainable agricultural development.

[0003] Currently, the assessment of bee conservation effectiveness mainly relies on two mainstream technologies: traditional molecular marker methods and conventional morphological determination methods. However, both of these methods have significant limitations in practical applications and are difficult to meet the needs of efficient, low-cost, and high-precision conservation monitoring. Summary of the Invention

[0004] The purpose of this application is to address the shortcomings of the prior art by providing a method, apparatus, electronic device, and storage medium for evaluating the effectiveness of bee conservation, thereby improving the accuracy of bee conservation evaluation.

[0005] To achieve the above objectives, the technical solutions adopted in the embodiments of this application are as follows: In a first aspect, embodiments of this application provide a method for evaluating the effectiveness of bee conservation, the method comprising: Obtain multiple original morphological indicators of the target bee species; Multiple original morphological indicators of the bees to be tested are input into a pre-trained XGBoost model for learning and analysis, and the XGBoost model outputs multiple target morphological indicators corresponding to the target species. The bee species to be tested are evaluated based on the target morphological indicators to obtain the evaluation results of the preservation effect of the bees to be tested.

[0006] Optionally, the target morphological indicators corresponding to the target variety include: snout length, sixth ventral plate width, sixth ventral plate length, forewing vein angle D7, forewing vein angle A4, forewing vein angle N23, forewing vein angle 026, third ventral plate wax mirror distance, cubital vein length, and forewing vein angle L13.

[0007] Optionally, the parameters used during training of the XGBoost model include at least the following: a total of 100 weak learners; a learning rate of 0.3; a minimum loss reduction required for tree splitting of 0; a maximum tree depth of 15; a minimum sum of the Hessian matrices in the child nodes of 1; a sample ratio of 1 used when training each tree; a feature ratio of 0.8 used when constructing each tree; a binary logistic regression objective function; and a log loss as the evaluation criterion.

[0008] Optionally, the step of evaluating the species of the bee to be tested based on each of the target morphological indicators to obtain the target evaluation result includes: Each of the target morphological indicators is input into a pre-constructed classification and recognition model, which then determines the conservation effect evaluation value of the bees under test based on each target morphological indicator, thus obtaining the target evaluation result.

[0009] Optionally, the training process of the XGBoost model is as follows: Multiple sample data were collected, including sample data of the target bee species and sample data of the comparison bee species. Each sample data included all the original morphological indicators obtained in advance. Missing values ​​in the multiple sample data are filled in to obtain multiple processed sample data. The first proportion of the processed multiple sample data is used as the training set, and the second proportion of the processed multiple sample data is used as the test set. The training set is input into the initial XGBoost model for training, and the trained XGBoost model is validated based on the test set to obtain the XGBoost model.

[0010] Optionally, the step of outputting at least one target morphological index corresponding to the bee under test by the pre-trained XGBoost model includes: The pre-trained XGBoost model outputs scores for each original morphological index based on the original morphological index. The scores of each of the aforementioned morphological indicators are sorted to obtain the sorted original morphological indicators. Based on the sorted original morphological indicators, the multiple target morphological indicators are determined and output.

[0011] Optionally, determining and outputting the plurality of target morphological indicators based on the sorted original morphological indicators includes: Select a predetermined number of original morphological indicators from the sorted original morphological indicators as the multiple target morphological indicators.

[0012] Secondly, embodiments of this application also provide an evaluation device for the effectiveness of bee conservation, the device comprising: The acquisition module is used to acquire multiple original morphological indicators of the target bee species. The learning module is used to input multiple original morphological indicators of the bees to be tested into a pre-trained XGBoost model for learning and analysis, and the XGBoost model outputs multiple target morphological indicators corresponding to the target species. The determination module is used to evaluate the species of the bee to be tested based on each of the target morphological indicators, and to obtain the evaluation result of the preservation effect of the bee to be tested.

[0013] Optionally, the target morphological indicators corresponding to the target variety include: snout length, sixth ventral plate width, sixth ventral plate length, forewing vein angle D7, forewing vein angle A4, forewing vein angle N23, forewing vein angle 026, third ventral plate wax mirror distance, cubital vein length, and forewing vein angle L13.

[0014] Optionally, the parameters used during training of the XGBoost model include at least the following: a total of 100 weak learners; a learning rate of 0.3; a minimum loss reduction required for tree splitting of 0; a maximum tree depth of 15; a minimum sum of the Hessian matrices in the child nodes of 1; a sample ratio of 1 used when training each tree; a feature ratio of 0.8 used when constructing each tree; a binary logistic regression objective function; and a log loss as the evaluation criterion.

[0015] Optionally, the determining module is specifically used for: Each of the target morphological indicators is input into a pre-constructed classification and recognition model, which then determines the conservation effect evaluation value of the bees under test based on each target morphological indicator, thus obtaining the target evaluation result.

[0016] Optionally, the training module is used for: Multiple sample data were collected, including sample data of the target bee species and sample data of the comparison bee species. Each sample data included all the original morphological indicators obtained in advance. Missing values ​​in the multiple sample data are filled in to obtain multiple processed sample data. The first proportion of the processed multiple sample data is used as the training set, and the second proportion of the processed multiple sample data is used as the test set. The training set is input into the initial XGBoost model for training, and the trained XGBoost model is validated based on the test set to obtain the XGBoost model.

[0017] Optionally, the learning module is specifically used for: The pre-trained XGBoost model outputs scores for each original morphological index based on the original morphological index. The scores of each of the aforementioned morphological indicators are sorted to obtain the sorted original morphological indicators. Based on the sorted original morphological indicators, the multiple target morphological indicators are determined and output.

[0018] Optionally, the learning module is specifically used for: Select a predetermined number of original morphological indicators from the sorted original morphological indicators as the multiple target morphological indicators.

[0019] Thirdly, embodiments of this application also provide an electronic device, including: a processor, a memory, and a bus. The memory stores program instructions executable by the processor. When the application runs, the processor communicates with the memory via the bus, and the processor executes the program instructions to perform the steps of the bee conservation effect evaluation method described in the first aspect above.

[0020] Fourthly, embodiments of this application also provide a computer-readable storage medium storing a computer program, which is read and executes the steps of the bee conservation effectiveness evaluation method described in the first aspect.

[0021] The beneficial effects of this application are: This application provides a method, apparatus, electronic device, and storage medium for evaluating the conservation effect of honeybees. The method involves acquiring multiple raw morphological indicators of the target honeybee species; inputting these indicators into a pre-trained XGBoost model for learning and analysis; and having the XGBoost model output multiple target morphological indicators corresponding to the target species. The method then evaluates the honeybee species based on these target morphological indicators to obtain the conservation effect assessment results. The XGBoost model standardizes the intelligent screening and evaluation process for key morphological indicators, avoiding human bias and improving analytical efficiency. It addresses the pain points of existing technologies, such as high cost, low efficiency, and strong subjectivity. Rapid detection is achieved, significantly improving the efficiency and scientific rigor of conservation efforts. Attached Figure Description

[0022] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 A flowchart illustrating a method for evaluating the effectiveness of bee conservation provided in an embodiment of this application; Figure 2 A flowchart illustrating a model training method provided in an embodiment of this application; Figure 3 A schematic diagram of an apparatus for evaluating the effectiveness of bee conservation provided in an embodiment of this application; Figure 4 This is a structural block diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the accompanying drawings in this application are for illustrative and descriptive purposes only and are not intended to limit the scope of protection of this application. Furthermore, it should be understood that the schematic drawings are not drawn to scale. The flowcharts used in this application illustrate operations implemented according to some embodiments of this application. It should be understood that the operations in the flowcharts may not be implemented in sequence, and steps without logical contextual relationships may be reversed or implemented simultaneously. In addition, those skilled in the art, guided by the content of this application, may add one or more other operations to the flowcharts, or remove one or more operations from the flowcharts.

[0025] Furthermore, the described embodiments are merely some, not all, of the embodiments of this application. The components of the embodiments of this application described and illustrated herein can typically be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0026] It should be noted that the term "comprising" will be used in the embodiments of this application to indicate the presence of the features declared thereafter, but does not exclude the addition of other features.

[0027] Optionally, the remote sensing image display method provided in this application embodiment can be applied to an electronic device, such as a mobile phone, tablet computer, laptop computer, PDA, desktop computer, or other terminal device with computing and display capabilities, or it can be a server. Specifically, it can be applied to applications in terminal devices, such as mobile phone applications (APP) and computer application systems.

[0028] The following section will explain in detail the specific implementation process of evaluating the bee conservation effect provided in the embodiments of this application.

[0029] Figure 1 This is a flowchart illustrating a method for evaluating the effectiveness of bee conservation, as provided in an embodiment of this application. The subject executing this method is the aforementioned electronic device. Figure 1 As shown, the method includes: S101. Obtain multiple original morphological indicators of the target bee species.

[0030] The target species is the Hunchun black bee, which exhibits strong cold resistance, disease resistance, and ecological adaptability. These multiple primitive morphological indicators refer to a series of anatomical characteristic parameters obtained through microscopic measurement or image analysis of the bees under test. All primitive morphological indicators were collected according to internationally accepted bee morphology measurement standards to ensure the accuracy and comparability of the data.

[0031] Optionally, after the bee to be tested has fully extended its proboscis, it is preserved in alcohol, and all primitive morphological indicators of the bee are measured completely. These primitive morphological indicators can be, for example, 40 primitive morphological indicators, specifically including key traits of multiple body regions such as head indicators, thorax indicators, abdomen indicators, and wing indicators, thereby constructing a dataset that comprehensively reflects the phenotypic characteristics of bees. Among them, head indicators may include morphological indicators such as head width, head length, distance between compound eyes, and antennae length; thorax indicators may include thorax width, thorax length, forewing length and width, and hindwing-related indicators; abdomen indicators may include the length and width of each abdominal segment, and wax mirror-related dimensions; wing indicators may include specific angles such as forewing vein angle; other key indicators may include proboscis length, cubital vein length, and distance between wax mirrors of the third abdominal septum.

[0032] S102. Input multiple original morphological indicators of the bees to be tested into the pre-trained XGBoost model for learning and analysis, and output multiple target morphological indicators corresponding to the target species by the XGBoost model.

[0033] Among them, the number of multiple target morphological indicators is less than the number of multiple original morphological indicators, that is, multiple target morphological indicators belong to a portion of multiple original morphological indicators.

[0034] Optionally, after obtaining all the original morphological indicators of the bees to be tested, these indicators are used as input variables into a pre-trained XGBoost model for feature importance evaluation and key morphological indicator selection. The XGBoost model is a highly efficient ensemble learning algorithm based on gradient boosting decision trees. Due to its good generalization ability, anti-overfitting properties, and advantages in handling high-dimensional sparse data, the XGBoost model is widely used in bioinformatics and phenomics. The model file for this XGBoost model includes the trained tree structure, weight parameters, and feature processing rules.

[0035] Specifically, before inputting multiple raw morphological indicators into the XGBoost model, the raw morphological indicators need to be standardized to ensure that the feature order is completely consistent with that during model training. Then, the standardized data is input into the XGBoost model, and the XGBoost model's corresponding inference interface is called to obtain the importance score of each raw morphological indicator for the classification of Hunchun black bees. Finally, multiple target morphological indicators are determined and output according to the importance scores of each raw morphological indicator.

[0036] S103. Evaluate the bee species to be tested based on each target morphological index to obtain the evaluation results of the preservation effect of the bees to be tested.

[0037] Optionally, after obtaining multiple key target morphological indicators, the evaluation phase begins. Step S103 aims to synthesize these most representative target morphological indicators to quantitatively determine whether the bee species being tested conforms to the standard phenotypic characteristics of the target bee species, thereby evaluating its conservation effect.

[0038] Specifically, if all the target morphological indicators of the bee under test fall within the typical range of Hunchun black bee values, the bee is assessed as having "high purity" or "belonging to purebred". Conversely, if they do not, the bee is assessed as having "hybridization risk" or "deviation from the standard phenotype". The assessment results can be expressed not only as category labels, such as "qualified" or "unqualified", but also as continuous probability values, such as a purity score between 0 and 1, which facilitates tiered management and dynamic monitoring by beekeeping managers.

[0039] In this embodiment, multiple raw morphological indicators of the target bee species are acquired; these indicators are then input into a pre-trained XGBoost model for learning and analysis. The XGBoost model outputs multiple target morphological indicators corresponding to the target species; and the bees are evaluated based on these indicators to obtain an assessment of their conservation effectiveness. The XGBoost model standardizes the intelligent screening and evaluation process for key morphological indicators, avoiding human bias and improving analytical efficiency. This addresses the pain points of existing technologies, such as high cost, low efficiency, and strong subjectivity. Rapid on-site detection is achieved, significantly improving the efficiency and scientific rigor of conservation efforts.

[0040] Optionally, the target morphological indicators corresponding to the above-mentioned target varieties include: proboscis length, sixth abdominal plate width, sixth abdominal plate length, forewing vein angle D7, forewing vein angle A4, forewing vein angle N23, forewing vein angle 026, third abdominal plate wax mirror distance, cubital vein length, and forewing vein angle. That is, the target morphological indicators for Hunchun black bees include the aforementioned 10 target morphological indicators, which can best distinguish Hunchun black bees from other bee species in terms of key morphological indicators.

[0041] Among them, the proboscis length can be represented by Proboscis_length, the width of the sixth sternite can be represented by Width_of_sternite_6_(T6), the length of the sixth sternite can be represented by Length_of_sternite_6_(L6), the forewing vein angle D7 can be represented by Forewing_vein_angle_D7, the forewing vein angle A4 can be represented by Forewing_vein_angle_A4, the forewing vein angle N23 can be represented by Forewing_vein_angle_N23, the forewing vein angle 026 can be represented by Forewing_vein_angle_026, the distance between the wax mirrors of the third sternite can be represented by Distance_between_wax_mirrors_on_sternite_3, the cubital vein length can be represented by Cubital_vein_a_(a), and the forewing vein angle L13 can be represented by Forewing_vein_angle_L13.

[0042] Optionally, the parameters used during training of the above XGBoost model include at least the following: a total of 100 weak learners; a learning rate of 0.3; a minimum loss reduction required for tree splitting of 0; a maximum tree depth of 15; a minimum sum of the Hessian matrices in the child nodes of 1; a sample ratio of 1 used when training each tree; a feature ratio of 0.8 used when constructing each tree; a binary logistic regression objective function; and a log loss as the evaluation criterion.

[0043] In this context, a weak learner represents an ensemble model consisting of 100 regression trees. The learner values ​​are chosen to ensure the model's expressive power while avoiding the waste of computational resources and the risk of overfitting caused by excessive iteration. In this embodiment, when the number of weak learners exceeds 100, the model's AUC value on the test set tends to stabilize, and adding more base learners has limited performance improvement.

[0044] The learning rate controls the weight of each tree's contribution to the final prediction result. A higher learning rate allows for faster convergence in the early stages, and combined with a sufficient number of iterations, it helps improve training efficiency. In this embodiment, 0.3 has been verified as the optimal learning rate that balances speed and accuracy.

[0045] The minimum loss reduction required for tree splitting indicates that node splitting is only allowed when the loss function reduction after splitting is greater than or equal to the minimum loss reduction required for tree splitting. Setting the minimum loss reduction required for tree splitting to 0 means that there are no restrictions on splitting conditions, allowing the model to explore more potential structures. This is suitable for the characteristics of complex nonlinear relationships between morphological indicators in this embodiment, and is beneficial for capturing subtle but critical discrimination patterns.

[0046] The maximum depth of a tree represents the limit on the maximum level depth of each decision tree. Deeper trees can model more complex feature interactions, and are particularly suitable for handling higher-order combination effects between multiple morphological indicators in bees, such as implicit rules like "proboscis length × forewing vein angle". Although deep trees have a slight tendency to overfit, this has been effectively suppressed through other regularization mechanisms, such as subsampling and column sampling.

[0047] The minimum sum of the Hessian matrices in the child nodes can be used to control the lower bound of the sum of the Hessian gradients of the samples in the leaf nodes. Setting it to 1 indicates that even if a leaf node contains only a very small number of samples or has weak gradient information, its existence is allowed. This enhances the model's sensitivity to rare phenotypic patterns, making it particularly suitable for identifying marginal individuals that may appear in conservation populations.

[0048] The proportion of samples used when training each tree is the same as the proportion of all training samples used each time a weak learner is built. Since the dataset used in this embodiment is of high quality, has no significant noise, and has a balanced sample distribution, there is no need to introduce additional randomness to enhance robustness. On the contrary, it can improve the stability and reproducibility of the model. Therefore, the proportion of samples used when training each tree is set to 1.

[0049] Setting the feature ratio used when constructing each tree to 80% means that 80% of the original morphological indicators are randomly selected as candidate splitting features when generating each tree. This sampling strategy effectively reduces the correlation between trees, increases the diversity of the ensemble model, thereby enhancing the overall generalization ability and preventing the model from over-relying on a few strong features, which could lead to amplified bias.

[0050] The objective function is a binary logistic regression, indicating that the goal of model training is to determine whether the bee being tested belongs to the target bee species, such as the Hunchun black bee, which is a typical binary classification problem. Logistic regression is chosen as the objective function, and the output is the probability value of belonging to the positive class, facilitating subsequent threshold setting and risk classification management.

[0051] The evaluation criterion is log loss, which is used as a performance monitoring metric during model training. This metric comprehensively considers classification accuracy and prediction confidence, effectively reflecting the model's ability to distinguish uncertain samples, and is particularly suitable for evaluation scenarios that require quantifying "seed purity".

[0052] Optionally, the evaluation of the bee under test based on each target morphological index in S103 above, to obtain the target evaluation result, may include: Each target morphological index is input into a pre-built classification and recognition model. The classification and recognition model determines the conservation effect assessment value of the bees under test based on each target morphological index, and the target assessment result is obtained.

[0053] Optionally, in step S102, the XGBoost model is mainly used for feature importance ranking and key morphological index screening, its core function being to reduce the dimensionality and focus the high-dimensional original data. The classification and recognition model used in step S103 undertakes the final discrimination task, that is, using a small number of carefully selected target morphological indicators as input to determine species affiliation or to quantitatively assess the purity of the species. For example, multiple target morphological indicators obtained in step S102 can be input into the classification and recognition model, and the classification and recognition module can output a predicted probability value for whether the bee in question belongs to a purebred based on each target morphological indicator.

[0054] For example, if the output predicted probability value is greater than or equal to 0.9, the bee under test can be determined to have high purity and be a qualified breeding bee; if the output predicted probability is between 0.6 and 0.9, the bee under test is determined to have medium purity and can be kept in isolation; if the output predicted probability is less than 0.6, the bee under test is determined to have low purity or be a hybrid individual and should be culled or its breeding restricted.

[0055] Figure 2 This is a flowchart illustrating a model training method provided in an embodiment of this application, as shown below. Figure 2 As shown, the training process of the XGBoost model is as follows: S201. Collect data from multiple samples.

[0056] The data from multiple samples can include data from the target bee species and data from comparison bee species. Each sample includes all pre-measured original morphological indicators. The target species could be, for example, the Hunchun Black Bee, while comparison bee species could be, for example, the Italian Bee or the Caucasian Bee. For each bee of each species, precise measurements can be taken under a microscope according to universal bee morphology standards to obtain all original morphological indicators for each bee.

[0057] S202. Complete the missing values ​​in the multiple sample data to obtain the processed multiple sample data. Use the first proportion of the processed multiple sample data as the training set and the second proportion of the processed multiple sample data as the test set.

[0058] Optionally, since some data items may be missing due to operational errors, sample damage, or equipment failure during actual measurement, it is necessary to clean and complete the original morphological indicators obtained from the measurement to improve the stability and reliability of model training. For example, the mean imputation method, K-nearest neighbor imputation method, regression prediction imputation method, and the method of deleting extremely missing samples can be used for imputation processing.

[0059] The first ratio can be 60%, the second ratio can be 40%, or other ratio values, depending on actual needs.

[0060] S203. Input the training set into the initial XGBoost model for training, and validate the trained XGBoost model based on the test set to obtain the XGBoost model.

[0061] Specifically, the sample data from the training set can be input into the initial XGBoost model to initiate the gradient boosting iterative process. Each new tree uses the residual from the previous round as its learning target, gradually approaching the optimal classification boundary. During training, the trend of the loss function can be monitored, and an early stopping mechanism can be enabled. That is, training is automatically terminated when the loss no longer decreases after 10 consecutive rounds of validation to prevent overfitting. Then, the performance of the trained model is validated using sample data from the test set. Once the validation is successful, the XGBoost model is obtained.

[0062] Optionally, in S102 above, the output of at least one target morphological index corresponding to the bee under test by the pre-trained XGBoost model may include: the pre-trained XGBoost model outputting the scores of each original morphological index based on each original morphological index; sorting the scores of each morphological index to obtain sorted original morphological indices; and determining and outputting multiple target morphological indices based on the sorted original morphological indices.

[0063] The sorting can be done from high to low or from low to high. If sorted from high to low, the top 10 pattern indicators can be used as the target pattern indicators; if sorted from low to high, the bottom 10 pattern indicators can be used as the target pattern indicators.

[0064] Optionally, the above process of determining and outputting multiple target morphological indicators based on the sorted original morphological indicators may include: Select a preset number of original morphological indicators from the sorted original morphological indicators as multiple target morphological indicators.

[0065] Figure 3 A schematic diagram of an apparatus for evaluating the effectiveness of bee conservation provided in an embodiment of this application is shown below. Figure 3 As shown, the device includes: The acquisition module 301 is used to acquire multiple original morphological indicators of the target bee species. The learning module 302 is used to input multiple original morphological indicators of the bee under test into the pre-trained XGBoost model for learning and analysis, and the XGBoost model outputs multiple target morphological indicators corresponding to the target species. The determination module 303 is used to evaluate the bees to be tested based on each of the target morphological indicators, and to obtain the evaluation result of the preservation effect of the bees to be tested.

[0066] Optionally, the target morphological indicators corresponding to the target variety include: snout length, sixth ventral plate width, sixth ventral plate length, forewing vein angle D7, forewing vein angle A4, forewing vein angle N23, forewing vein angle 026, third ventral plate wax mirror distance, cubital vein length, and forewing vein angle L13.

[0067] Optionally, the parameters used during training of the XGBoost model include at least the following: a total of 100 weak learners; a learning rate of 0.3; a minimum loss reduction required for tree splitting of 0; a maximum tree depth of 15; a minimum sum of the Hessian matrices in the child nodes of 1; a sample ratio of 1 used when training each tree; a feature ratio of 0.8 used when constructing each tree; a binary logistic regression objective function; and a log loss as the evaluation criterion.

[0068] Optionally, the determining module 303 is specifically used for: Each of the target morphological indicators is input into a pre-constructed classification and recognition model, which then determines the conservation effect evaluation value of the bees under test based on each target morphological indicator, thus obtaining the target evaluation result.

[0069] Optionally, training module 304 is used for: Multiple sample data were collected, including sample data of the target bee species and sample data of the comparison bee species. Each sample data included all the original morphological indicators obtained in advance. Missing values ​​in the multiple sample data are filled in to obtain multiple processed sample data. The first proportion of the processed multiple sample data is used as the training set, and the second proportion of the processed multiple sample data is used as the test set. The training set is input into the initial XGBoost model for training, and the trained XGBoost model is validated based on the test set to obtain the XGBoost model.

[0070] Optionally, the learning module 302 is specifically used for: The pre-trained XGBoost model outputs scores for each original morphological index based on the original morphological index. The scores of each of the aforementioned morphological indicators are sorted to obtain the sorted original morphological indicators. Based on the sorted original morphological indicators, the multiple target morphological indicators are determined and output.

[0071] Optionally, the learning module 302 is specifically used for: Select a predetermined number of original morphological indicators from the sorted original morphological indicators as the multiple target morphological indicators.

[0072] Figure 4 This is a structural block diagram of an electronic device 400 provided in an embodiment of this application. (See diagram below.) Figure 4 As shown, the electronic device may include: a processor 401 and a memory 402.

[0073] Optionally, a bus 403 may also be included, wherein the memory 402 is used to store machine-readable instructions executable by the processor 401. When the electronic device 400 is running, the processor 401 and the memory 402 communicate with each other via the bus 403, and the processor 401 executes the machine-readable instructions to perform the method steps in the above method embodiments.

[0074] This application also provides a computer-readable storage medium storing a computer program, which, when run by a processor, executes the method steps described in the above-described method for evaluating the effectiveness of bee conservation.

[0075] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and devices described above can be referred to the corresponding processes in the method embodiments, and will not be repeated here. In the several embodiments provided in this application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed mutual coupling or direct coupling or communication connection can be through some communication interfaces; the indirect coupling or communication connection of devices or modules can be electrical, mechanical, or other forms.

[0076] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. If the functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes: USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, optical disks, and other media capable of storing program code.

[0077] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for evaluating the effectiveness of bee conservation, characterized in that, The method includes: Obtain multiple original morphological indicators of the target bee species; Multiple original morphological indicators of the bees to be tested are input into a pre-trained XGBoost model for learning and analysis, and the XGBoost model outputs multiple target morphological indicators corresponding to the target species. The bee species to be tested are evaluated based on the target morphological indicators to obtain the evaluation results of the preservation effect of the bees to be tested.

2. The method for evaluating the effectiveness of bee conservation according to claim 1, characterized in that, The target morphological indicators corresponding to the target variety include: snout length, sixth ventral plate width, sixth ventral plate length, forewing vein angle D7, forewing vein angle A4, forewing vein angle N23, forewing vein angle 026, third ventral plate wax mirror distance, cubital vein length, and forewing vein angle L13.

3. The method for evaluating the effectiveness of bee conservation according to claim 1, characterized in that, The XGBoost model uses at least the following parameters during training: a total of 100 weak learners; a learning rate of 0.3; a minimum loss reduction required for tree splitting of 0; a maximum tree depth of 15; a minimum sum of the Hessian matrices in the child nodes of 1; a sample ratio of 1 used to train each tree; a feature ratio of 0.8 used to construct each tree; a binary logistic regression objective function; and a log loss as the evaluation criterion.

4. The method for evaluating the effectiveness of bee conservation according to claim 1, characterized in that, The evaluation of the bee species to be tested based on each of the target morphological indicators to obtain the target evaluation results includes: Each of the target morphological indicators is input into a pre-constructed classification and recognition model, which then determines the conservation effect evaluation value of the bees under test based on each target morphological indicator, thus obtaining the target evaluation result.

5. The method for evaluating the effectiveness of bee conservation according to claim 1, characterized in that, The training process of the XGBoost model is as follows: Multiple sample data were collected, including sample data of the target bee species and sample data of the comparison bee species. Each sample data included all the original morphological indicators obtained in advance. Missing values ​​in the multiple sample data are filled in to obtain multiple processed sample data. The first proportion of the processed multiple sample data is used as the training set, and the second proportion of the processed multiple sample data is used as the test set. The training set is input into the initial XGBoost model for training, and the trained XGBoost model is validated based on the test set to obtain the XGBoost model.

6. The method for evaluating the effectiveness of bee conservation according to claim 1, characterized in that, The output of at least one target morphological index corresponding to the bee under test by the pre-trained XGBoost model includes: The pre-trained XGBoost model outputs scores for each original morphological index based on the original morphological index. The scores of each of the aforementioned morphological indicators are sorted to obtain the sorted original morphological indicators. Based on the sorted original morphological indicators, the multiple target morphological indicators are determined and output.

7. The method for evaluating the effectiveness of bee conservation according to claim 6, characterized in that, The step of determining and outputting the multiple target morphological indicators based on the sorted original morphological indicators includes: Select a predetermined number of original morphological indicators from the sorted original morphological indicators as the multiple target morphological indicators.

8. A device for evaluating the effectiveness of bee conservation, characterized in that, include: The acquisition module is used to acquire multiple original morphological indicators of the target bee species. The learning module is used to input multiple original morphological indicators of the bees to be tested into a pre-trained XGBoost model for learning and analysis, and the XGBoost model outputs multiple target morphological indicators corresponding to the target species. The determination module is used to evaluate the species of the bee to be tested based on each of the target morphological indicators, and to obtain the evaluation result of the preservation effect of the bee to be tested.

9. An electronic device, characterized in that, The device includes a memory and a processor, the memory storing a computer program executable by the processor, the processor executing the computer program to implement the steps of the method for evaluating the effectiveness of bee conservation as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, performs the steps of the method for evaluating the effectiveness of bee conservation as described in any one of claims 1-7.

Images