Digital rural information processing method and system

Abstract

The invention discloses a digital rural information processing method and system, and relates to the technical field of agricultural data analysis, and the method comprises the steps: collecting digital rural agricultural data through a sensor, extracting agricultural data features, and employing information gain, correlation analysis and recursive feature elimination to screen the agricultural data features, an XGBoost model is constructed, model optimization is carried out through a tree structured Padow estimation method, and the optimized XGBoost model is adopted to analyze agricultural data features; selecting agricultural decision features by using an SHAP value method in combination with an XGBoost model, generating a rural agricultural decision by using an adversarial generative network according to an analysis result of the XGBoost model and the agricultural decision features, and performing decision optimization; and displaying and storing the rural agricultural decision in a rural database. According to the invention, the representativeness of agricultural data features is improved, the accuracy of agricultural data analysis is improved, automatic adjustment of rural agriculture is realized, and the precision of agricultural management decision is further optimized.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural data analysis, and particularly to a digital rural information processing method and system. Background Art

[0002] With the rapid development of digital technologies, the construction of digital villages has become an important trend in the development of modern agriculture. The core of digital villages is to collect, process, and analyze agricultural data through technical means such as the Internet of Things, sensors, and artificial intelligence to achieve precise agricultural management and decision-making. During the agricultural production process, various sensors are used to obtain various types of data such as meteorology, soil, and crop growth, and integrate them into a digital platform to provide a scientific basis for agricultural management. In recent years, with the development of technologies such as big data, cloud computing, and artificial intelligence, the application of digital villages has gradually expanded to all aspects of agricultural production, including crop planting, resource management, agricultural decision-making, and other fields.

[0003] Existing agricultural data processing technologies generally adopt standard feature selection and machine learning models. Although they can solve some basic agricultural prediction problems, there are still many deficiencies. The existing technologies fail to fully explore the potential associations in agricultural data, resulting in limited prediction accuracy and generalization ability of the models, and often neglect the dynamic optimization of the agricultural decision-making process, lacking a real-time feedback and decision adjustment mechanism for changes in agricultural data. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a digital rural information processing method and system to solve the problems that the existing technologies fail to fully explore the potential associations in agricultural data, resulting in limited prediction accuracy and generalization ability of the models, and often neglect the dynamic optimization of the agricultural decision-making process.

[0006] To solve the above technical problems, the present invention provides the following technical solutions:

[0007] In the first aspect, the present invention provides a digital rural information processing method, which includes

[0008] collecting digital rural agricultural data through sensors, extracting agricultural data features and screening agricultural data features by using information gain, correlation analysis, and recursive feature elimination, constructing an XGBoost model and optimizing the model by using the tree-structured Parzen estimator method, and analyzing agricultural data features by using the optimized XGBoost model; using the SHAP value method in combination with the XGBoost model to select agricultural decision-making features, and generating rural agricultural decisions by using a generative adversarial network according to the analysis results of the XGBoost model and agricultural decision-making features and performing decision optimization; displaying the rural agricultural decisions and storing them in a rural database.

[0009] As a preferred solution of the digital rural information processing method described in the present invention, wherein: extracting agricultural data features, screening agricultural data features by using information gain, correlation analysis and recursive feature elimination, constructing an XGBoost model and optimizing the model by using the tree-structured Parzen estimation method, and analyzing agricultural data features by using the optimized XGBoost model includes

[0010] Using feature engineering to extract agricultural data features from rural agricultural data, forming a feature set N with the extracted agricultural data features, calculating the information gain of each feature in the feature set N, and screening the agricultural data features with information gain greater than a preset threshold to form a preliminary screening set;

[0011] Randomly forming feature pairs in the preliminary screening set, calculating the Pearson correlation coefficient of the feature pairs, and marking the feature pairs with Pearson correlation coefficient greater than a set correlation threshold as similar feature pairs;

[0012] Training a linear regression model by using the recursive feature elimination method based on all similar feature pairs, calculating the weight of each feature in the similar feature pairs, sorting them from largest to smallest to form a feature sequence, constructing an empty secondary screening set S, removing the feature with the smallest weight and moving the similar features of the removed feature from the feature sequence to the secondary screening set S, and repeating the use of the recursive feature elimination method based on the removed feature sequence until the number of features in the secondary screening set S reaches a preset value and then stopping;

[0013] Constructing an XGBoost model includes an error term and a regularization term ;

[0014] Using the gradient boosting algorithm to construct an error update objective function q;

[0015] Defining a hyperparameter space according to the XGBoost model, and calculating the conditional probability of the hyperparameters by using the tree-structured Parzen estimation method ;

[0016] Calculating the hyperparameter expectation through the expected improvement strategy ;

[0017] Iterating the hyperparameters through the gradient boosting algorithm, selecting hyperparameters for XGBoost model training in each iteration, calculating the hyperparameter expectation of each iteration, synchronously evaluating the XGBoost model by using cross-validation, and stopping the iteration when the hyperparameter expectation converges, and selecting the hyperparameter with the largest expectation as the optimal hyperparameter of the XGBoost model;

[0018] Inputting the secondary screening set S into the trained XGBoost model to obtain an agricultural analysis result.

[0019] As a preferred solution of the digital rural information processing method described in the present invention, wherein: the method of using the SHAP value method in combination with the XGBoost model to select agricultural decision-making features, and using the adversarial generation network to generate rural agricultural decisions and optimize decisions according to the analysis results of the XGBoost model and agricultural decision-making features means to select feature i from the feature set, and add feature i to the secondary screening set S, and calculate the feature contribution using the SHAP value method. ;

[0020] If the feature contribution i of feature i is greater than 0, then add feature i to the secondary screening set S, traverse all features in the feature set to calculate the feature contribution, and add all qualified features to the secondary screening set. Based on the added secondary screening set and the corresponding XGBoost model analysis results, construct an adversarial generation network for training, and generate rural agricultural decisions through the adversarial generation network;

[0021] Construct a convolutional neural network to evaluate the adaptability of rural agricultural decisions, and output the rural agricultural decisions that pass the adaptability evaluation.

[0022] As a preferred solution of the digital rural information processing method described in the present invention, wherein: after generating the rural agricultural decision, it is evaluated by relevant departments and professionals. After passing the evaluation, the implementation means to display the output rural agricultural decision to relevant departments and professionals, judge the decision feasibility through comprehensive evaluation by professional scoring, and implement the rural agricultural decision evaluated by professionals.

[0023] As a preferred solution of the digital rural information processing method described in the present invention, wherein: the collection of digital rural agricultural data through sensors means to use soil humidity sensors, meteorological sensors and video monitoring in rural agricultural areas to collect rural agricultural data and preprocess it to unify the format.

[0024] As a preferred solution of the digital rural information processing method described in the present invention, wherein: the display of rural agricultural decisions means to display the implemented rural agricultural decisions in real time, and collect rural agricultural data through sensors to display the implementation progress of rural agricultural decisions.

[0025] As a preferred solution of the digital rural information processing method described in the present invention, wherein: the storage in the rural database means to form records of rural agricultural decisions and rural agricultural data collected during the implementation process and store them in the rural database. The rural database classifies and stores the records with timestamps attached and uploads them to the cloud for backup.

[0026] In the second aspect, the present invention provides a digital rural information processing system, including,

[0027] A data acquisition module for collecting digital rural agricultural data through sensors and performing preprocessing;

[0028] A data analysis module for extracting agricultural data features, screening agricultural data features, constructing an XGBoost model, optimizing the model through the tree-structured Parzen estimation method, and analyzing agricultural data features using the optimized XGBoost model;

[0029] A decision generation module for using the SHAP value method to combine with the XGBoost model to select agricultural decision features, and generating rural agricultural decisions using an adversarial generative network based on the analysis results of the XGBoost model and agricultural decision features and performing decision optimization;

[0030] A display and storage module for displaying rural agricultural decisions and storing them in a rural database.

[0031] In a third aspect, the present invention provides a computer device, including a memory and a processor, where the memory stores a computer program, and: when the computer program is executed by the processor, any step of the digital rural information processing method described in the first aspect of the present invention is implemented.

[0032] In a fourth aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored, and: when the computer program is executed by the processor, any step of the digital rural information processing method described in the first aspect of the present invention is implemented.

[0033] The beneficial effects of the present invention are as follows: The present invention collects rural agricultural data and uses information gain, correlation analysis, and recursive feature elimination to screen agricultural data features, improving the representativeness of agricultural data features, and optimizing the analysis of agricultural data features by optimizing the XGBoost model through the tree-structured Parzen estimation method, improving the accuracy of agricultural data analysis. And generating rural agricultural decisions using an adversarial generative network based on the analysis results of the XGBoost model and agricultural decision features realizes the automatic adjustment of rural agriculture, further optimizing the accuracy of agricultural management decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0035] Figure 1 It is a flowchart of the digital rural information processing method in Embodiment 1.

[0036] Figure 2This is a structural diagram of the digital village information processing system in Example 1. DETAILED DESCRIPTION

[0037] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0038] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0039] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0040] Example 1, reference Figure 1 and Figure 2 , which is the first embodiment of the present invention, and provides a digital village information processing method, comprising the following steps:

[0041] S1. Collect digital rural agricultural data through sensors, extract agricultural data features and use information gain, correlation analysis and recursive feature elimination to screen agricultural data features, build an XGBoost model and optimize the model through the tree structured Pazen estimation method, and use the optimized XGBoost model to analyze agricultural data features;

[0042] Specifically, collecting digital rural agricultural data through sensors refers to using soil moisture sensors, meteorological sensors, and video surveillance in rural agricultural areas to collect rural agricultural data and pre-processing and then unifying the data into a unified format.

[0043] Furthermore, agricultural data features are extracted and information gain, correlation analysis and recursive feature elimination are used to screen agricultural data features. XGBoost model is constructed and optimized by tree-structured Pazen estimation method. The optimized XGBoost model is used to analyze agricultural data features, including

[0044] Feature engineering is used to extract agricultural data features from rural agricultural data, and the extracted agricultural data features are formed into a feature set N. The information gain of each feature in the feature set N is calculated, and the agricultural data features with information gain greater than a preset threshold are screened to form a preliminary screening set.

[0045] Randomly form feature pairs in the preliminary screening set, calculate the Pearson correlation coefficient of the feature pairs, and mark the feature pairs with a Pearson correlation coefficient greater than the set correlation threshold as similar feature pairs;

[0046] Based on all similar feature pairs, train a linear regression model using the recursive feature elimination method, calculate the weight of each feature in the similar feature pairs, sort them from largest to smallest to form a feature sequence, construct an empty secondary screening set S, remove the feature with the smallest weight, and move the similar features of the removed feature from the feature sequence to the secondary screening set S. Repeat using the recursive feature elimination method based on the removed feature sequence until the number of features in the secondary screening set S reaches the preset value and then stop;

[0047] Construct the XGBoost model including the error term and the regularization term :

[0048]

[0049] where is the i-th true value, is the i-th predicted value, is the number of samples, is the total number of leaf nodes of the tree in the XGBoost model, is the weight of the j-th leaf node, and are hyperparameters that control the tree complexity in the XGBoost model;

[0050] Use the gradient boosting algorithm to construct the error update objective function q:

[0051]

[0052] where is the value of the error update objective function at the t-th iteration, is the value of the error update objective function at the (t - 1)-th iteration, is the learning rate, is the error of the XGBoost model at the t-th iteration;

[0053] Define the hyperparameter space according to the XGBoost model, and use the tree-structured Parzen estimation method to calculate the conditional probability of the hyperparameters :

[0054]

[0055] where is the probability distribution of good samples (i.e., hyperparameters with better performance), is the probability distribution of bad samples (i.e., hyperparameters with poor performance), obtained from historical data, is the preset value of the objective function;

[0056] Calculate the hyperparameter expectation through the expected improvement strategy :

[0057]

[0058] Iterate the hyperparameters through the gradient boosting algorithm. In each iteration, select hyperparameters for XGBoost model training, calculate the hyperparameter expectation of each iteration, synchronously evaluate the XGBoost model using cross-validation, and stop the iteration when the hyperparameter expectation converges. Select the hyperparameter with the maximum expectation as the optimal hyperparameter of the XGBoost model;

[0059] Input the secondary screening set S into the trained XGBoost model to obtain the agricultural analysis result.

[0060] By combining techniques such as information gain, Pearson correlation coefficient, and recursive feature elimination, a multi-level feature screening process is formed. Compared with traditional feature selection methods, this step can accurately select features strongly correlated with the target variable and eliminate redundant or low-correlated features, thus significantly improving the prediction accuracy and computational efficiency of the model. Through this refined feature selection method, it can be ensured that the model can effectively utilize the most predictive data features and avoid the negative impact of feature redundancy on the model performance. By using the tree-structured Parzen estimator (TPE) method to optimize the hyperparameters of the XGBoost model, the present invention overcomes the problem of the large amount of computational resources required by traditional hyperparameter optimization methods (such as grid search). TPE can dynamically adjust the distribution of hyperparameters according to historical data, reduce the ineffective search area, and ensure the efficiency and accuracy of the optimization process. Through the expected improvement strategy, the model can quickly find the optimal hyperparameter combination, thereby improving the overall performance of the model. Especially when dealing with complex agricultural data, it can significantly improve the prediction accuracy and stability of the model. Through the agricultural analysis results generated by the optimized XGBoost model, combined with agricultural data characteristics for decision optimization, the intelligent level of agricultural management is further improved.

[0061] S2. Use the SHAP value method to combine with the XGBoost model to select agricultural decision-making features, and use the adversarial generative network to generate rural agricultural decisions and optimize the decisions according to the XGBoost model analysis results and agricultural decision-making features;

[0062] Specifically, the SHAP value method is used in combination with the XGBoost model to select agricultural decision-making features. According to the analysis results of the XGBoost model and the agricultural decision-making features, a generative adversarial network is used to generate rural agricultural decisions and optimize the decisions, which means to select feature i from the feature set and add feature i to the secondary screening set S, and calculate the feature contribution using the SHAP value method. :

[0063]

[0064] Where i is the feature added to the secondary screening set S. is the analysis result of the XGBoost model using the secondary screening set S. is the analysis result of the XGBoost model using the secondary screening set after adding feature i.

[0065] If the feature contribution i of feature i is greater than 0, then feature i is added to the secondary screening set S. Traverse all features in the feature set to calculate the feature contribution, and add all qualified features to the secondary screening set. Based on the added secondary screening set and the corresponding XGBoost model analysis results, a generative adversarial network is constructed for training, and rural agricultural decisions are generated through the generative adversarial network.

[0066] A convolutional neural network is constructed to conduct an adaptability assessment of the rural agricultural decisions, and the rural agricultural decisions that pass the adaptability assessment are output.

[0067] By introducing the SHAP value method, the contribution of each feature to the prediction result of the XGBoost model can be accurately calculated. This method can quantify the actual influence of each feature in the model, avoid the disadvantages of blind screening, and thus improve the interpretability of the model and the accuracy of decision-making. By traversing all features in the feature set and calculating their feature contributions, the present invention can not only accurately identify the features with important influences, but also construct a more refined secondary screening set S by adding these features. This process of feature selection helps to remove redundant or irrelevant features, avoid the overfitting problem caused by feature redundancy, and thus improve the generalization ability of the agricultural data analysis model. The application of GAN enables the generation of agricultural decisions not only to rely on historical data, but also to combine real-time analysis results, with stronger adaptability. The decisions generated by GAN can be optimized in different agricultural production environments, providing personalized decision-making suggestions, thus enhancing the scientificity and practicality of agricultural decisions. Through actual adaptability assessment by CNN, it is ensured that the generated decisions meet the actual needs of agricultural production. CNN can deeply analyze all aspects of agricultural decisions through its powerful feature extraction ability, so as to judge whether the decision adapts to the changes in the actual agricultural environment.

[0068] Furthermore, after generating the rural agricultural decision, it is evaluated by relevant departments and professionals. After passing the evaluation, the implementation means presenting the output rural agricultural decision to relevant departments and professionals, and judging the feasibility of the decision through comprehensive evaluation by professionals' scoring. The rural agricultural decision that has passed the evaluation by professionals is implemented.

[0069] By submitting the generated rural agricultural decision to relevant departments and professionals for evaluation, the decision does not solely rely on the analysis results of the automated model, but combines the perspective of actual operation. In agricultural decision-making, although machine learning models and algorithms can provide valuable analysis, they often lack a deep understanding of specific agricultural environments and local differences. Through the evaluation of professionals, these local differences and potential problems can be better captured, ensuring that the decision-making plan has a high degree of practical feasibility. Professionals can analyze from multiple dimensions by scoring the rural agricultural decision, such as economic benefits, environmental impact, social acceptance, etc. This multi-dimensional scoring mechanism is more comprehensive than the traditional single-standard evaluation and can comprehensively consider all aspects of the decision. Each scoring criterion will be customized according to actual agricultural production needs, so as to ensure that the decision-making plan can be fully verified at all levels.

[0070] S3. Present the rural agricultural decision and store it in the rural database.

[0071] Specifically, collecting digital rural agricultural data through sensors means using soil moisture sensors, meteorological sensors, and video surveillance in rural agricultural areas to collect rural agricultural data and preprocess it into a unified format.

[0072] Furthermore, presenting the rural agricultural decision means presenting the implemented rural agricultural decision in real time and collecting rural agricultural data through sensors to show the implementation progress of the rural agricultural decision.

[0073] This embodiment also provides a digital rural information processing system, including:

[0074] A data acquisition module, used to collect digital rural agricultural data through sensors and preprocess it;

[0075] A data analysis module, used to extract agricultural data features, screen agricultural data features, construct an XGBoost model, optimize the model through the tree-structured Parzen estimation method, and analyze agricultural data features using the optimized XGBoost model;

[0076] A decision generation module, used to select agricultural decision features by combining the XGBoost model using the SHAP value method, and generate a rural agricultural decision using a generative adversarial network according to the analysis results of the XGBoost model and agricultural decision features and perform decision optimization;

[0077] A display storage module is used to display rural agricultural decisions and store them in the rural database.

[0078] This embodiment also provides a computer device applicable to the digital rural information processing method, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the digital rural information processing method proposed in the above embodiment.

[0079] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a carrier network, NFC (Near Field Communication), or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a button, a trackball, or a touchpad provided on the housing of the computer device, or an external keyboard, a touchpad, or a mouse, etc.

[0080] This embodiment also provides a storage medium on which a computer program is stored, and when the program is executed by a processor, it implements the digital rural information processing method proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic memory, flash memory, a magnetic disk, or an optical disc.

[0081] In summary, the present invention collects rural agricultural data and uses information gain, correlation analysis, and recursive feature elimination to screen agricultural data features, improving the representativeness of agricultural data features. The tree-structured Parzen estimation method is used to optimize the XGBoost model for analyzing agricultural data features, improving the accuracy of agricultural data analysis. Based on the analysis results of the XGBoost model and agricultural decision-making features, a generative adversarial network is used to generate rural agricultural decisions, realizing the automatic adjustment of rural agriculture and further optimizing the accuracy of agricultural management decisions.

[0082] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A digital village information processing method, characterized in that: include, Collect digital rural agricultural data through sensors, extract agricultural data features and use information gain, correlation analysis and recursive feature elimination to screen agricultural data features, build an XGBoost model and optimize the model through the tree-structured Pazen estimation method, and use the optimized XGBoost model to analyze agricultural data features; The SHAP value method was combined with the XGBoost model to select agricultural decision features, and the adversarial generative network was used to generate rural agricultural decisions and optimize the decisions based on the XGBoost model analysis results and agricultural decision features. Display and store rural agricultural decisions in the rural database.

2. The digital village information processing method according to claim 1, characterized in that: The agricultural data features are extracted and the agricultural data features are screened by using information gain, correlation analysis and recursive feature elimination, an XGBoost model is constructed and the model is optimized by using a tree-structured Pazen estimation method, and the optimized XGBoost model is used to analyze the agricultural data features. Feature engineering is used to extract agricultural data features from rural agricultural data, and the extracted agricultural data features are formed into a feature set N. The information gain of each feature in the feature set N is calculated, and the agricultural data features with information gain greater than a preset threshold are screened to form a preliminary screening set. Randomly form feature pairs in the preliminary screening set, calculate the Pearson correlation coefficient of the feature pairs, and mark the feature pairs whose Pearson correlation coefficient is greater than a set correlation threshold as similar feature pairs; Based on all similar feature pairs, a linear regression model is trained using a recursive feature elimination method, and the weight of each feature in the similar feature pairs is calculated. The features are sorted from large to small to form a feature sequence, and an empty secondary screening set S is constructed. The feature with the smallest weight is removed and the similar features of the removed features are moved from the feature sequence to the secondary screening set S. The recursive feature elimination method is repeatedly used based on the removed feature sequence until the number of features in the secondary screening set S reaches a preset value and stops; Building the XGBoost model including the error term and the regularization term ; Use the gradient boosting algorithm to construct the error update objective function q; Define the hyperparameter space according to the XGBoost model and use the tree-structured Patzen estimation method to calculate the conditional probability of the hyperparameters ; Calculating hyperparameter expectations using the expected improvement strategy ; Iterate the hyperparameters through the gradient boosting algorithm, select hyperparameters for XGBoost model training in each iteration, calculate the hyperparameter expectations for each iteration, and simultaneously use cross-validation to evaluate the XGBoost model. When the hyperparameter expectations converge, stop the iteration and select the hyperparameters with the largest expectations as the optimal hyperparameters of the XGBoost model. The secondary screening set S is input into the trained XGBoost model to obtain the agricultural analysis results.

3. The digital village information processing method according to claim 2, characterized in that: The method of using the SHAP value method in combination with the XGBoost model to select agricultural decision features, and using the adversarial generative network to generate rural agricultural decisions and optimize the decision based on the XGBoost model analysis results and agricultural decision features, refers to selecting feature i from the feature set, adding feature i to the secondary screening set S, and using the SHAP value method to calculate the feature contribution ; If the feature contribution i of feature i is greater than 0, then feature i is added to the secondary screening set S, all features in the feature set are traversed to calculate the feature contribution, and all the features that meet the conditions are added to the secondary screening set. Based on the added secondary screening set and the corresponding XGBoost model analysis results, a generative adversarial network is constructed for training, and rural agricultural decisions are generated through the generative adversarial network; Construct a convolutional neural network to conduct adaptive assessment of rural agricultural decisions, and output the rural agricultural decisions that have passed the adaptive assessment.

4. The digital village information processing method according to claim 3, characterized in that: After the rural agricultural decision is generated, it will be evaluated by relevant departments and professionals. Implementation after evaluation means that the output rural agricultural decision will be displayed to relevant departments and professionals, and the feasibility of the decision will be judged after comprehensive evaluation by professionals. The rural agricultural decision evaluated by professionals will be implemented.

5. The digital village information processing method according to claim 1, characterized in that: The collection of digital rural agricultural data through sensors refers to the use of soil moisture sensors, meteorological sensors and video surveillance in rural agricultural areas to collect rural agricultural data and pre-process the data into a unified format.

6. The digital village information processing method according to claim 5, characterized in that: The display of rural agricultural decisions refers to displaying the implemented rural agricultural decisions in real time, and collecting rural agricultural data through sensors to display the implementation progress of the rural agricultural decisions.

7. The digital village information processing method according to claim 6, characterized in that: The storage in the rural database refers to recording the rural agricultural data collected during the rural agricultural decision-making and implementation process in the rural database, which attaches timestamps to the records, stores them in classified categories, and uploads them to the cloud for backup.

8. A digital village information processing system, based on the digital village information processing method according to any one of claims 1 to 7, characterized in that: include, Data acquisition module, used to collect digital rural agricultural data through sensors and perform pre-processing; The data analysis module is used to extract and screen agricultural data features, build an XGBoost model and optimize the model through the tree-structured Pazen estimation method, and use the optimized XGBoost model to analyze agricultural data features; The decision generation module is used to select agricultural decision features using the SHAP value method combined with the XGBoost model, and to generate rural agricultural decisions and optimize them using the adversarial generative network based on the XGBoost model analysis results and agricultural decision features; The display and storage module is used to display and store rural agricultural decisions in the rural database.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the digital village information processing method described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the digital village information processing method described in any one of claims 1 to 7 are implemented.

Images