Low-income population identification method, system and device based on neural network, and medium
Through the neural network-based identification method, the problem of large manual workload and inconvenient economic development levels in traditional low-income population identification methods is solved, and the rapid and accurate identification of low-income population is achieved, adapting to the differences in economic development levels in different regions is improved. The identification accuracy rate is improved.
Patent Information
- Application Number
- CN202510144302.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-10
- Publication Date
- 2025-05-16
AI Technical Summary
The traditional low-income population identification method has a large manual workload and is not suitable for the differences in economic development levels in different regions, resulting in unfair and inefficient identification.
Using a neural network-based identification method, through data extraction, information screening, data cleaning, feature coding, data set division and neural network model construction and training, a fully connected neural network model is built to present a pyramid shape, feature extraction is performed, and website publishing and generation services are built through FLASK.
It has achieved rapid and accurate identification of low-income populations, adapted to differences in economic development levels in different regions, improved identification accuracy, and provided a basis for policy adjustments.
Smart Images
Figure CN120011886A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of data analysis, and in particular to a method, system, device and medium for identifying low-income population based on a neural network. Background Art
[0002] In terms of identifying low-income population, the traditional review method is mainly based on reference documents of relevant specifications, and low-income population is divided through manual statistics and review.
[0003] The traditional review method involves a huge amount of manual work, and the identification of low-income people has regional characteristics. The economic development levels of different regions are inconsistent. If a unified judgment is made through reference documents, it will be unfair to low-income people in need of assistance in economically developed regions.
[0004] Therefore, how to quickly and efficiently identify low-income workers while taking into account the differences caused by the inconsistency of economic development levels in different regions is a technical problem that needs to be solved urgently. Summary of the invention
[0005] The technical task of the present invention is to provide a low-income population identification method, system, device and medium based on neural network to solve the problem of how to quickly and efficiently identify low-income people by sight, while taking into account the differences caused by the inconsistency of economic development levels in different regions.
[0006] The technical task of the present invention is achieved in the following way: a method for identifying low-income people based on a neural network, the method is specifically as follows:
[0007] Data extraction and information screening: Extract relevant information about the low-income, extremely poor and non-protected population from the low-income dynamic monitoring platform, and screen out the district name, age, marital status, academic status, labor capacity and serious disease information of the low-income, extremely poor and non-protected population;
[0008] Data cleaning: Process missing values, outliers, and duplicate values for the filtered information, and extract the low-income identification field to construct the predicted value Y;
[0009] Feature encoding: Label Encoding and One-Hot Encoding are used to encode the extracted non-digital coded feature values to obtain the encoded features, and the encoded features are standardized to obtain standardized data; among them, the non-digital coded feature values include district names, health conditions and educational levels;
[0010] Dataset division: The extracted low-income identification field is used to construct the predicted value Y and the standardized data to form a data set, and the data set is divided into a training set and a test set;
[0011] Build a neural network model: Build a fully connected neural network model by stacking dense layers, make the fully connected neural network model appear in a pyramid shape, and perform feature extraction to capture useful combination features and obtain corresponding feature representations;
[0012] Neural network training: The neural network model is trained through the training set and tested through the test set. After several iterations of testing, the preset recognition accuracy is achieved.
[0013] Create an interface: Use FLASK to build a website and publish the generated service.
[0014] Preferably, the method further includes feature selection and feature extraction, as follows:
[0015] Perform feature analysis on the data set and use heat maps to display the impact of each field on the final data, filter out effective feature fields, and obtain better display effects.
[0016] As a preference, the way to handle missing values depends on the importance of the field. For the missing age, age is closely related to labor capacity and income. Taking the median and mode will affect the determination of low-income population, so the missing age entry is deleted;
[0017] For serious diseases, if there is a missing value, it is assumed to be no disease.
[0018] Preferably, when building a neural network model, the number of network layers, the number of neurons in the fully connected layer, and the hyperparameters of the relevant neural network model are set.
[0019] Preferably, build a website through FLASK to publish the generated services as follows:
[0020] Build network nodes through FLASK;
[0021] Initialize the weights of the neural network model before requesting;
[0022] The parameters are passed to the neural network model, passed through the neural network model layer by layer, and finally the corresponding discrimination category is calculated and returned through the interface.
[0023] A low-income population identification system based on a neural network, the system comprising:
[0024] The data extraction and information screening module is used to extract relevant information of the low-income, extremely poor and non-protected population from the low-income dynamic monitoring platform, and screen out the district name, age, marital status, academic status, labor capacity and serious disease information of the low-income, extremely poor and non-protected population;
[0025] The data cleaning module is used to process missing values, outliers and duplicate values of the filtered information, and extract the low-income identification field to construct the predicted value Y;
[0026] The feature encoding module is used to perform feature encoding using Label Encoding and One-Hot Encoding for the extracted non-digital coded feature values, obtain the encoded features, and perform standardization on the encoded features to obtain standardized data; wherein the non-digital coded feature values include district names, health conditions, and educational levels;
[0027] A data set partitioning module is used to construct a prediction value Y and standardized data from the extracted low-income identification field to form a data set, and to partition the data set into a training set and a test set;
[0028] The neural network model building module is used to build a fully connected neural network model by stacking dense layers, construct the fully connected neural network model in a pyramid shape, and perform feature extraction to capture useful combined features and obtain corresponding feature representations;
[0029] The neural network training module is used to train the neural network model through the training set and test the neural network model through the test set. Finally, after several iterative tests, the preset recognition accuracy is achieved;
[0030] Interface creation module, used to build website publishing generation services through FLASK.
[0031] Preferably, the system also includes a feature selection and feature extraction module, which is used to perform feature analysis on the data set, and display the impact of each field on the final data in the form of a heat map, filter out effective feature fields, and obtain better display effects.
[0032] Preferably, the interface creation module includes:
[0033] Node building submodule, used to build network nodes through FLASK;
[0034] The initialization submodule is used to initialize the weights of the neural network model before requesting;
[0035] The result return submodule is used to pass parameters to the neural network model, pass them layer by layer through the neural network model, and finally calculate the corresponding discrimination category and return it through the interface.
[0036] An electronic device comprising: a memory and at least one processor;
[0037] Wherein, the memory stores a computer program;
[0038] The at least one processor executes the computer program stored in the memory, so that the at least one processor performs the low-income population identification method based on neural network as described above.
[0039] A computer-readable storage medium having a computer program stored therein, wherein the computer program can be executed by a processor to implement the neural network-based low-income population identification method as described above.
[0040] The low-income population identification method, system, device and medium based on neural network of the present invention have the following advantages:
[0041] (1) The present invention adopts a data-driven approach and a strategy of identifying low-income people based on a training neural network model. This takes into account the differences caused by the inconsistency of different economic development levels. It only needs to input the basic information of the judge to achieve the identification of low-income people. It is easy to operate, and the predicted results can provide a basis for the adjustment of later policies. At the same time, the accuracy of low-income entry recognition is improved;
[0042] (ii) After the data-driven training is completed, the weight data file of the corresponding model is saved, a website node is built through Flask, and it is published through a proxy server in the form of an interface for other systems to call. BRIEF DESCRIPTION OF THE DRAWINGS
[0043] The present invention is further described below in conjunction with the accompanying drawings.
[0044] Attached Figure 1 It is a flowchart of the low-income population identification method based on neural network;
[0045] Attached Figure 2 Screenshot of the debugging process interface for the network optimal solution;
[0046] Attached Figure 3 It is a schematic diagram of the parameter information of the neural network model;
[0047] Attached Figure 4 Take a screenshot of the neural network model training results interface;
[0048] Attached Figure 5 A schematic diagram of the training loss and training accuracy changes during the training of the neural network model. DETAILED DESCRIPTION
[0049] The neural network-based low-income population identification method, system, device and medium of the present invention are described in detail below with reference to the drawings and specific embodiments of the specification.
[0050] Embodiment 1:
[0051] As attached Figure 1 As shown, this embodiment provides a low-income population identification method based on a neural network, and the method is specifically as follows:
[0052] S1. Data extraction and information screening: Extract relevant information of the low-income, extremely poor and non-protected population from the low-income dynamic monitoring platform, and screen out the district name, age, marital status, academic status, labor capacity and serious disease information of the low-income, extremely poor and non-protected population;
[0053] S2, data cleaning: missing values, outliers and duplicate values are processed for the screened information, and the low-income identification field is extracted to construct the predicted value Y;
[0054] S3. Feature encoding: Label Encoding and One-Hot Encoding are used to encode the extracted non-digital coded feature values, obtain the encoded features, and standardize the encoded features to obtain standardized data; the non-digital coded feature values include district names, health status and educational level; in the process of feature encoding, it is necessary to constrain the size of the value, which is conducive to determining that the impact value of low-income people should be larger, and the fields that have no significant correlation or cannot be specifically measured need to be scaled to a smaller state, such as the district name, which is a string of characters, also needs to be re-encoded;
[0055] S4, data set division: construct the predicted value Y and the standardized data from the extracted low-income identification field to form a data set, and divide the data set into a training set and a test set;
[0056] S5. Build a neural network model: Build a fully connected neural network model by stacking dense layers, make the fully connected neural network model appear in a pyramid shape, and perform feature extraction to capture useful combination features and obtain corresponding feature representations;
[0057] S6, neural network training: The neural network model is trained through the training set, and the neural network model is tested through the test set. Finally, after several iterative tests, the preset recognition accuracy is achieved;
[0058] S7. Create interface: Use FLASK to build a website to publish and generate services.
[0059] This embodiment also includes feature selection and feature extraction, which are as follows:
[0060] Perform feature analysis on the data set and use heat maps to display the impact of each field on the final data, filter out effective feature fields, and obtain better display effects.
[0061] In step S2 of this embodiment, the way of processing missing values depends on the importance of the field. For the missing age, age is closely related to labor capacity and income. Both the median and the mode will affect the determination of low-income population, so the missing age entry is deleted;
[0062] For serious diseases, if there is a missing value, it is assumed to be no disease.
[0063] When building a neural network model in step S5 of this embodiment, the number of network layers is set, the number of neurons in the fully connected layer is set, and the hyperparameters of the relevant neural network model are set.
[0064] In step S6 of this embodiment, the website publishing and generating service is built by FLASK as follows:
[0065] S601, build network nodes through FLASK;
[0066] S602, before requesting, initializing the weights of the neural network model;
[0067] S603, passing the parameters to the neural network model, passing them layer by layer through the neural network model, and finally calculating the corresponding discrimination category and returning it through the interface.
[0068] As attached Figure 2 As shown in the figure, the structure of the network uses a feature pyramid to extract features. The number of neurons in each layer decreases layer by layer. The optimizer uses SGD to optimize the network weights. The learning rate is set to 0.001, and regularization is used to prevent overfitting.
[0069] Considering the authenticity of the data, the model is trained with 70% of the data from the real low-income population database obtained from the dynamic monitoring platform of a certain region, and the remaining 30% of the data is used as test data. After processing the data, some training results of the neural network are shown in the attached figure. Figure 3 As shown;
[0070] The training loss during model training and the change in training accuracy are shown in the attached figure. Figure 4 As shown, the horizontal axis represents the round of training, the vertical axis on the left represents the training loss, and the vertical axis on the right represents the training accuracy.
[0071] Finally, after testing the model on the test set, the accuracy of the model can reach 86.31%.
[0072] Embodiment 2:
[0073] This embodiment provides a low-income population identification system based on a neural network, the system comprising:
[0074] The data extraction and information screening module is used to extract relevant information of the low-income, extremely poor and non-protected population from the low-income dynamic monitoring platform, and screen out the district name, age, marital status, academic status, labor capacity and serious disease information of the low-income, extremely poor and non-protected population;
[0075] The data cleaning module is used to process missing values, outliers and duplicate values of the filtered information, and extract the low-income identification field to construct the predicted value Y;
[0076] The feature encoding module is used to perform feature encoding using Label Encoding and One-Hot Encoding for the extracted non-digital coded feature values, obtain the encoded features, and perform standardization on the encoded features to obtain standardized data; wherein the non-digital coded feature values include district names, health conditions, and educational levels;
[0077] A data set partitioning module is used to construct a prediction value Y and standardized data from the extracted low-income identification field to form a data set, and to partition the data set into a training set and a test set;
[0078] The neural network model building module is used to build a fully connected neural network model by stacking dense layers, construct the fully connected neural network model in a pyramid shape, and perform feature extraction to capture useful combined features and obtain corresponding feature representations;
[0079] The neural network training module is used to train the neural network model through the training set and test the neural network model through the test set. Finally, after several iterative tests, the preset recognition accuracy is achieved;
[0080] Interface creation module, used to build website publishing generation services through FLASK.
[0081] This embodiment also includes a feature selection and feature extraction module, which is used to perform feature analysis on the data set, and display the impact of each field on the final data in the form of a heat map, filter out effective feature fields, and obtain better display effects.
[0082] The interface creation module in this embodiment includes:
[0083] Node building submodule, used to build network nodes through FLASK;
[0084] The initialization submodule is used to initialize the weights of the neural network model before requesting;
[0085] The result return submodule is used to pass parameters to the neural network model, pass them layer by layer through the neural network model, and finally calculate the corresponding discrimination category and return it through the interface.
[0086] Embodiment 3:
[0087] This embodiment also provides an electronic device, including: a memory and at least one processor;
[0088] Wherein, the memory stores computer-executable instructions;
[0089] The at least one processor executes the computer-executable instructions stored in the memory, so that the at least one processor executes the low-income population identification method based on neural network in any embodiment of the present invention.
[0090] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor may be a microprocessor or any conventional processor, etc.
[0091] The memory can be used to store computer programs and / or modules. The processor realizes various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function, etc.; the data storage area can store data created according to the use of the terminal, etc. In addition, the memory can also include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (SMC), a secure digital (SD) card, a flash memory card, at least one disk storage period, a flash memory device, or other volatile solid-state storage devices.
[0092] Embodiment 4:
[0093] This embodiment also provides a computer-readable storage medium, in which a plurality of instructions are stored, and the instructions are loaded by a processor, so that the processor executes the low-income population identification method based on a neural network in any embodiment of the present invention. Specifically, a system or device equipped with a storage medium can be provided, on which a software program code that implements the functions of any of the above embodiments is stored, and a computer (or CPU or MPU) of the system or device reads and executes the program code stored in the storage medium.
[0094] In this case, the program code itself read from the storage medium can realize the function of any one of the above-mentioned embodiments, and thus the program code and the storage medium storing the program code constitute a part of the present invention.
[0095] The storage medium embodiments for providing the program code include a floppy disk, a hard disk, a magneto-optical disk, an optical disk (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RYM, DVD-RW, DVD+RW), a magnetic tape, a non-volatile memory card, and a ROM. Alternatively, the program code can be downloaded from a server computer via a communication network.
[0096] In addition, it should be clear that the functions of any of the above embodiments can be implemented not only by executing the program code read by the computer, but also by enabling an operating system operating on the computer to complete part or all of the actual operations based on instructions from the program code.
[0097] In addition, it can be understood that the program code read from the storage medium is written to a memory provided in an expansion board inserted into the computer or written to a memory provided in an expansion unit connected to the computer, and then based on the instructions of the program code, a CPU installed on the expansion board or the expansion unit is enabled to perform part or all of the actual operations, thereby realizing the functions of any of the above-mentioned embodiments.
[0098] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for identifying low-income population based on neural network, characterized in that: The method is as follows: Data extraction and information screening: Extract relevant information about the low-income, extremely poor and non-protected population from the low-income dynamic monitoring platform, and screen out the district name, age, marital status, academic status, labor capacity and serious disease information of the low-income, extremely poor and non-protected population; Data cleaning: Process missing values, outliers, and duplicate values for the filtered information, and extract the low-income identification field to construct the predicted value Y; Feature encoding: Label Encoding and One-Hot Encoding are used to encode the extracted non-digital coded feature values to obtain the encoded features, and the encoded features are standardized to obtain standardized data; among them, the non-digital coded feature values include district names, health conditions, and educational levels; Dataset division: The extracted low-income identification field is used to construct the predicted value Y and the standardized data to form a dataset, and the dataset is divided into a training set and a test set; Build a neural network model: Build a fully connected neural network model by stacking dense layers, make the fully connected neural network model appear in a pyramid shape, and perform feature extraction to capture useful combination features and obtain corresponding feature representations; Neural network training: The neural network model is trained through the training set and tested through the test set. After several iterations of testing, the preset recognition accuracy is achieved. Create an interface: Use FLASK to build a website to publish and generate services.
2. The low-income population identification method based on neural network according to claim 1 is characterized in that: The method also includes feature selection and feature extraction, as follows: Perform feature analysis on the data set and use heat maps to display the impact of each field on the final data, filter out effective feature fields, and obtain better display effects.
3. The low-income population identification method based on neural network according to claim 1 is characterized in that: The way to handle missing values depends on the importance of the field. For missing age, age is closely related to labor capacity and income. Taking the median and mode will affect the determination of low-income population, so the missing age entry is deleted; For serious diseases, if there is a missing value, it is assumed to be no disease.
4. The low-income population identification method based on neural network according to claim 1 is characterized in that: When building a neural network model, set the number of network layers, the number of neurons in the fully connected layer, and the hyperparameters of the relevant neural network model.
5. The low-income population identification method based on a neural network according to any one of claims 1 to 4, characterized in that: The specific services for building a website and publishing it through FLASK are as follows: Build network nodes through FLASK; Initialize the weights of the neural network model before requesting; The parameters are passed to the neural network model, passed through the neural network model layer by layer, and finally the corresponding discrimination category is calculated and returned through the interface.
6. A low-income population identification system based on neural network, characterized in that: The system includes: The data extraction and information screening module is used to extract relevant information of the low-income, extremely poor and non-protected population from the low-income dynamic monitoring platform, and screen out the district name, age, marital status, academic status, labor capacity and serious disease information of the low-income, extremely poor and non-protected population; The data cleaning module is used to process missing values, outliers and duplicate values of the filtered information, and extract the low-income identification field to construct the predicted value Y; The feature encoding module is used to perform feature encoding using Label Encoding and One-Hot Encoding for the extracted non-digital coded feature values, obtain the encoded features, and perform standardization processing on the encoded features to obtain standardized data; wherein the non-digital coded feature values include district names, health conditions, and educational levels; A data set partitioning module is used to construct a prediction value Y and standardized data from the extracted low-income identification field to form a data set, and to partition the data set into a training set and a test set; The neural network model building module is used to build a fully connected neural network model by stacking dense layers, construct the fully connected neural network model in a pyramid shape, and perform feature extraction to capture useful combined features and obtain corresponding feature representations; The neural network training module is used to train the neural network model through the training set and test the neural network model through the test set. Finally, after several iterative tests, the preset recognition accuracy is achieved; Interface creation module, used to build website publishing generation services through FLASK.
7. The low-income population identification system based on neural network according to claim 6 is characterized in that: The system also includes feature selection and feature extraction modules, which are used to perform feature analysis on the data set, and display the impact of each field on the final data in the form of a heat map, filter out effective feature fields, and obtain better display effects.
8. The low-income population identification system based on neural network according to claim 6 or 7, characterized in that: The interface creation module includes: Node building submodule, used to build network nodes through FLASK; The initialization submodule is used to initialize the weights of the neural network model before requesting; The result return submodule is used to pass parameters to the neural network model, pass them layer by layer through the neural network model, and finally calculate the corresponding discrimination category and return it through the interface.
9. An electronic device, characterized in that: include: memory and at least one processor; Wherein, the memory stores a computer program; The at least one processor executes the computer program stored in the memory, so that the at least one processor performs the low-income population identification method based on neural network as described in any one of claims 1 to 5.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which can be executed by a processor to implement the low-income population identification method based on a neural network as described in any one of claims 1 to 5.