CN114708003B - An abnormal data detection method, device, equipment and readable storage medium - Google Patents

Abstract

The present invention relates to the field of data processing. Specifically, it relates to an abnormal data detection method, device, equipment and readable storage medium. The method obtains first information, and the first information is product sales data information of at least one store; Send the first information to the data preprocessing model to obtain the second information, which is the price characteristic data information and sales volume characteristic data information obtained by preprocessing the first information; send the second information Go to the anomaly detection model to detect abnormal data, and obtain the third information. The third information is the abnormal product sales data obtained by performing two clustering screenings on the second information; perform verification processing on the third information, Obtain the abnormal product sales data filtered by the model after verifying the parameters. This application integrates the advantages of two clustering algorithms, overcomes the shortcomings of the two algorithms, and achieves the effect of efficiently and accurately judging abnormal data.

Description

An abnormal data detection method, device, equipment and readable storage medium

Technical field

The present invention relates to the field of data processing, and specifically to an abnormal data detection method, device, equipment and readable storage medium.

Background technique

In recent years, Internet technology has developed rapidly, and the e-commerce industry has also embarked on the fast lane of development. "Online shopping" is becoming more and more popular among people because of its convenience, time saving and door-to-door delivery. As the scale of each platform continues to expand and the number of products continues to increase, some unfair business practices, such as false price markings and fraudulent orders, have also emerged, seriously violating e-commerce laws, and it is necessary to accurately identify such product data. For such a large number of products, if manual inspection and screening is performed, not only will the workload be huge, but there will also be omissions and errors. What is needed is a data detection method that can accurately locate abnormal products, reduce manual intervention costs and reduce error rates.

Contents of the invention

The purpose of the present invention is to provide an abnormal data detection method, device, equipment and readable storage medium to improve the above problems. In order to achieve the above objects, the technical solutions adopted by the present invention are as follows:

On the one hand, this application provides an abnormal data detection method, which method includes: obtaining first information, which is at least one store's merchandise sales data information; sending the first information to a data preprocessing model , obtain the second information, which is the price characteristic data information and sales volume characteristic data information obtained by preprocessing the first information; send the second information to the anomaly detection model for abnormal data detection, and obtain the third information, the third information is the abnormal product sales data obtained by performing two clustering screenings on the second information; the third information is sent to the verification module for processing, and the fourth information is obtained, and the fourth information The information is the abnormal product sales data filtered by the model after verifying the parameters.

In the second aspect, embodiments of the present application provide an abnormal data detection device, including:

A first acquisition unit, configured to acquire first information, which is merchandise sales data information of at least one store;

A first processing unit, configured to send the first information to a data preprocessing model to obtain second information, where the second information is price characteristic data information and sales volume characteristic data information obtained by preprocessing the first information;

The second processing unit is used to send the second information to the anomaly detection model for abnormal data detection, and obtain the third information. The third information is the abnormal products obtained by performing two clustering screenings on the second information. sales data;

The third processing unit is configured to send the third information to the verification module for processing to obtain fourth information. The fourth information is the abnormal product sales data screened by the model after verifying the parameters.

In a third aspect, embodiments of the present application provide an abnormal data detection device, where the device includes a memory and a processor. The memory is used to store the computer program; the processor is used to implement the steps of the abnormal data detection method when executing the computer program.

In a fourth aspect, embodiments of the present application provide a readable storage medium. A computer program is stored on the readable storage medium. When the computer program is executed by a processor, the steps of the above abnormal data detection method are implemented.

The beneficial effects of the present invention are:

By extracting the characteristics of product sales receipts and using two different clustering algorithms for secondary clustering, this application can accurately locate the product, reduce manual intervention and reduce the error rate, and the present invention uses a high-efficiency The clustering method processes the data for the first time, effectively reducing the amount of data that needs to be detected, and then uses a high-precision clustering method to process the first clustered data. This integrates the advantages of the two algorithms and overcomes the problem. The shortcomings of the two algorithms are eliminated to achieve the effect of efficiently and accurately judging abnormal data.

Additional features and advantages of the invention will be set forth in the description which follows, and in part will be apparent from the description, or may be learned by practice of embodiments of the invention. The objectives and other advantages of the invention may be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.

Description of the drawings

In order to explain the technical solutions of the embodiments of the present invention more clearly, the drawings required to be used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and therefore do not It should be regarded as a limitation of the scope. For those of ordinary skill in the art, other relevant drawings can be obtained based on these drawings without exerting creative efforts.

Figure 1 is a schematic flow chart of an abnormal data detection method described in the embodiment of the present invention;

Figure 2 is a schematic structural diagram of an abnormal data detection device described in the embodiment of the present invention;

Figure 3 is a schematic structural diagram of an abnormal data detection device described in the embodiment of the present invention.

Detailed ways

In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments These are some embodiments of the present invention, rather than all embodiments. The components of the embodiments of the invention generally described and illustrated in the figures herein may be arranged and designed in a variety of different configurations. Therefore, the following detailed description of the embodiments of the invention provided in the appended drawings is not intended to limit the scope of the claimed invention, but rather to represent selected embodiments of the invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts fall within the scope of protection of the present invention.

It should be noted that similar reference numerals and letters represent similar items in the following figures, therefore, once an item is defined in one figure, it does not need further definition and explanation in subsequent figures. Meanwhile, in the description of the present invention, the terms "first", "second", etc. are only used to differentiate the description and cannot be understood as indicating or implying relative importance.

Example 1

As shown in Figure 1, this embodiment provides an abnormal data detection method, which includes step S1, step S2, step S3 and step S4.

Step S1: Obtain first information, which is product sales data information of at least one store;

Step S2: Send the first information to the data preprocessing model to obtain second information. The second information is price characteristic data information and sales volume characteristic data information obtained by preprocessing the first information;

Step S3: Send the second information to the anomaly detection model for abnormal data detection to obtain third information. The third information is the abnormal product sales data obtained by performing two clustering screenings on the second information;

Step S4: Send the third information to the verification module for processing to obtain fourth information. The fourth information is the abnormal product sales data screened by the model after verifying the parameters.

It can be understood that the above abnormal data is abnormal data in product sales data.

It can be understood that by extracting the characteristics of product sales receipts and using two different clustering algorithms for secondary clustering, this application can accurately locate the product, reduce manual intervention and reduce the error rate, and the present invention uses A high-efficiency clustering method is used to process the data for the first time, effectively reducing the amount of data that needs to be detected, and then a high-precision clustering method is used to process the first clustered data, thus combining the advantages of the two algorithms. The integration overcomes the shortcomings of the two algorithms and achieves efficient and accurate judgment of abnormal data.

In a specific implementation manner of the present disclosure, step S2 includes step S21, step S22 and step S23.

Step S21: Perform data processing on the commodity sales data information, clear invalid data in the first information, and average-fill the incomplete data in the first information to obtain the first sub-information;

Step S22: Calculate the characteristic data of the first information based on the first sub-information and the preset calculation formula. The characteristic data of the first information includes price characteristic data and sales volume characteristic data;

Step S23: Normalize the feature data of the first information, and smooth the normalized feature data to obtain pre-processed first information.

It can be understood that this application preprocesses the product sales data, removes invalid data, and performs mean filling on the data. The filling method is to sum up the data corresponding to other months, and then calculate the average, and use the average as the incomplete data. Filling values, thereby reducing the error value generated during feature data extraction and increasing the accuracy of clustering.

In a specific implementation manner of the present disclosure, step S3 includes step S31, step S32 and step S33.

Step S31: Send the price characteristic data information in the second information to the first clustering module for clustering to obtain the first clustering information. The first clustering information is the abnormal data information in the price characteristic data information. ;

Step S32: Perform data correspondence mapping between the first clustering information and the sales volume characteristic data information in the second information to obtain second sub-information, where the second sub-information includes the sales volume characteristics corresponding to the first clustering information. Data information;

Step S33: Send the second sub-information to the second clustering module for processing to obtain second clustering information. The second clustering information is anomalies obtained by performing two clustering screenings on the second information. Product sales data.

It can be understood that the above steps are to perform a primary clustering of the price characteristic data information, and map the data after the first clustering with the sales volume characteristic data, select the data for the second clustering, and then perform a second clustering. In this way, the first clustering can efficiently filter the data, and the second clustering can only filter the data corresponding to the data filtered out in the first time. This can ensure that the advantages can be integrated while being efficient.

In a specific implementation manner of the present disclosure, step S31 includes step S311, step S312, step S313 and step S314.

Step S311: Traverse the price feature data information based on the preset first initial parameter information, and process the price feature data according to the cluster feature tree generation method in the BIRCH algorithm to obtain a price cluster feature tree;

Step S312: Obtain at least one clustering feature cluster based on the price clustering feature tree, and calculate the threshold range corresponding to each clustering feature cluster;

Step S313: Analyze all the threshold ranges, and use the smallest threshold range among all the threshold ranges as the normal threshold range for determining the normal point;

Step S314: Determine abnormal points in the price clustering feature tree based on the normal threshold range, and determine abnormal data information in the price feature data information based on the abnormal points.

It can be understood that the above steps parameterize the BIRCH algorithm with the preset first initial parameters, generate a clustering feature tree based on the price feature data information, and then cluster the price feature data information based on the clustering feature tree. class, get at least one cluster, then analyze the range of cluster sizes, select the smallest range as the threshold range of normal data, and then reversely determine the threshold range of abnormal data to obtain abnormal data, which can be efficient and fast The data is processed to reduce the amount of data that needs to be processed in the second clustering.

In a specific implementation manner of the present disclosure, step S33 includes step S331, step S332, step S333 and step S334.

Step S331: Perform data processing based on the preset second initial parameters and the data information in the second sub-information, where the data information in the second sub-information is converted into coordinate data points in the spatial coordinate system, based on For each of the coordinate data points, obtain the mutual reachable distance between each of the coordinate data points;

Step S332: Generate a weighted distance map based on the mutual reachability distance, and generate a minimum spanning tree of the mutual reachability distance based on the weighted distance map;

Step S333: Convert the minimum spanning tree into components of a hierarchical cluster structure according to the mutual reachability distance, and build a hierarchical cluster structure based on the components of the clustering hierarchy;

Step S334: Compress the hierarchical cluster structure, and classify the data information in the second sub-information based on the compressed hierarchical cluster structure to obtain abnormal data in the second sub-information.

It can be understood that the above steps set the parameters of the clustering algorithm by setting the second initial parameter, transform the second sub-information into spatial coordinates, and calculate the mutual reachability distance between each coordinate point, thereby increasing the algorithm's ability to deal with noise. Robustness, and then build a minimum spanning tree based on the mutual reachability distance, and then cluster the coordinate points to obtain the abnormal data in the second sub-information, so that the anomalies in each sales feature data information can be more accurately determined Data information.

In a specific implementation manner of the present disclosure, step S4 includes step S41, step S42, step S43, step S44, step S45 and step S46.

Step S41: Obtain third sub-information, which includes normal sales data information of historical commodities and abnormal sales data information of historical commodities;

Step S42: Divide the third sub-information into a test set and a verification set, and send the test set of the third sub-information to the second clustering module for processing to obtain historical abnormal product sales data;

Step S43: Compare the historical abnormal product sales data with the verification set to obtain verification result information;

Step S44: Perform gray correlation analysis on the verification result information and all initial parameters in the anomaly detection model to obtain the correlation between the verification result information and all the initial parameters;

Step S45: Adjust the initial parameters based on the verification result information and the correlation degree to obtain an anomaly detection model with adjusted initial parameters. If the verification result is that the test set is inconsistent with the verification set, adjust the The initial parameters with the greatest correlation between the verification results, until the verification results are consistent with the results of the test set and the verification set;

Step S46: Send the first information to the anomaly detection model after adjusting the initial parameters for a second anomaly detection, and filter the third information with the data after the second anomaly detection to obtain the screened anomalies. Product sales data.

It can be understood that the present invention classifies historical data, and then sends it to the anomaly detection model for detection based on the historical data, performs gray correlation analysis on the detection results and the initial parameters, and obtains the correlation between the detection results and the initial parameters, and The first initial parameter and the second initial parameter are adjusted based on the correlation degree to obtain an adjusted anomaly detection model, and then a secondary abnormal data screening is performed to obtain the screened abnormal product sales data.

Example 2

As shown in Figure 2, this embodiment provides an abnormal data detection device, which includes a first acquisition unit 701, a first processing unit 702, a second processing unit 703, and a third processing unit 704.

The first obtaining unit 701 is used to obtain first information, which is the product sales data information of at least one store;

The first processing unit 702 is used to send the first information to the data preprocessing model to obtain second information. The second information is price characteristic data information and sales volume characteristic data information obtained by preprocessing the first information. ;

The second processing unit 703 is configured to send the second information to the anomaly detection model for abnormal data detection to obtain third information. The third information is anomalies obtained by performing two clustering screenings on the second information. Product sales data;

The third processing unit 704 is configured to send the third information to the verification module for processing to obtain fourth information. The fourth information is the abnormal product sales data filtered by the model after verifying the parameters.

In a specific implementation of the present disclosure, the first processing unit 702 includes a first processing sub-unit 7021, a second processing sub-unit 7022 and a third processing sub-unit 7023.

The first processing subunit 7021 is used to perform data processing on the commodity sales data information, clear invalid data in the first information, and average-fill the incomplete data in the first information to obtain the first sub-information;

The second processing sub-unit 7022 is used to calculate the characteristic data of the first information based on the first sub-information and the preset calculation formula. The characteristic data of the first information includes price characteristic data and sales volume characteristic data. ;

The third processing sub-unit 7023 is used to normalize the feature data of the first information, and smooth the normalized feature data to obtain pre-processed first information.

In a specific implementation of the present disclosure, the second processing unit 703 includes a first clustering sub-unit 7031, a fourth processing sub-unit 7032 and a second clustering sub-unit 7033.

The first clustering subunit 7031 is used to send the price feature data information in the second information to the first clustering module for clustering to obtain the first clustering information, where the first clustering information is the price feature Abnormal data in the data;

The fourth processing subunit 7032 is used to perform data corresponding mapping between the first cluster information and the sales volume characteristic data information in the second information to obtain second sub-information, where the second sub-information includes the first cluster information. The sales volume characteristic data information corresponding to the category information;

The second clustering subunit 7033 is used to send the second sub-information to the second clustering module for processing to obtain the second clustering information. The second clustering information is performed by performing two operations on the second information. Abnormal product sales data obtained through sub-clustering screening.

In a specific implementation of the present disclosure, the first clustering subunit 7031 includes a third clustering subunit 70311, a fourth clustering subunit 70312, a fifth clustering subunit 70313, and a sixth clustering subunit. Unit 70314.

The third clustering subunit 70311 is used to traverse the price feature data information based on the preset first initial parameter information, and process the price feature data according to the cluster feature tree generation method in the BIRCH algorithm to obtain price clustering. Class feature tree;

The fourth clustering subunit 70312 is used to obtain at least one clustering feature cluster based on the price clustering feature tree, and obtain and calculate the threshold range corresponding to each clustering feature cluster;

The fifth clustering subunit 70313 is used to analyze all the threshold ranges, and use the minimum threshold range among all the threshold ranges as the normal threshold range for judging normal points;

The sixth clustering subunit 70314 is configured to determine abnormal points in the price cluster feature tree based on the normal threshold range, and determine abnormal data information in the price feature data information based on the abnormal points.

In a specific implementation of the present disclosure, the second clustering subunit 7033 includes a seventh clustering subunit 70331, an eighth clustering subunit 70332, a ninth clustering subunit 70333, and a tenth clustering subunit. Unit 70334.

The seventh clustering subunit 70331 is used to perform data processing based on the preset second initial parameters and the data information in the second sub-information, where the data information in the second sub-information is converted into a spatial coordinate system coordinate data points within, based on each of the coordinate data points, obtain the mutual reachable distance between each of the coordinate data points;

The eighth clustering subunit 70332 is configured to generate a weighted distance map based on the mutual reachability distance, and generate a minimum spanning tree of the mutual reachability distance based on the weighted distance map;

The ninth clustering subunit 70333 is used to convert the minimum spanning tree into components of a hierarchical cluster structure according to the mutual reachability distance, and build a hierarchical cluster structure based on the components of the clustering hierarchy;

The tenth clustering subunit 70334 is used to compress the hierarchical cluster structure and classify the data information in the second sub-information based on the compressed hierarchical cluster structure to obtain abnormal data in the second sub-information. .

In a specific implementation manner of the present disclosure, the third processing unit 704 includes a first acquisition sub-unit 7041, a fifth processing sub-unit 7042, a sixth processing sub-unit 7043, a seventh processing sub-unit 7044, an eighth processing Subunit 7045 and ninth processing subunit 7046.

The first acquisition sub-unit 7041 is used to acquire the third sub-information, which is the normal sales data information of historical commodities and the abnormal sales data information of historical commodities;

The fifth processing subunit 7042 is used to divide the third sub-information into a test set and a verification set, and send the test set of the third sub-information to the second clustering module for processing to obtain historical abnormal product sales. data;

The sixth processing subunit 7043 is used to compare the historical abnormal product sales data with the verification set to obtain verification result information;

The seventh processing subunit 7044 is used to perform gray correlation analysis on the verification result information and all the initial parameters in the anomaly detection model, and obtain the correlation degree between the verification result information and all the initial parameters;

The eighth processing subunit 7045 is used to adjust the initial parameters based on the verification result information and the correlation degree to obtain an anomaly detection model with adjusted initial parameters, wherein if the verification result is the test set and the verification set If inconsistent, adjust the initial parameter with the greatest correlation with the verification result until the verification result is that the test set and the verification set results are consistent;

The ninth processing subunit 7046 is used to send the first information to the anomaly detection model after adjusting the initial parameters for the second anomaly detection, and filter the third information with the data after the second anomaly detection. , to obtain the screened abnormal product sales data.

It should be noted that, regarding the device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be described in detail here.

Example 3

Corresponding to the above method embodiments, embodiments of the present disclosure also provide an abnormal data detection device. An abnormal data detection device described below and an abnormal data detection method described above may be mutually referenced.

FIG. 3 is a block diagram of an abnormal data detection device 800 according to an exemplary embodiment. As shown in Figure 3, the abnormal data detection device 800 may include: a processor 801 and a memory 802. The anomaly data detection device 800 may also include one or more of a multimedia component 803, an input/output (I/O) interface 804, and a communication component 805.

The processor 801 is used to control the overall operation of the abnormal data detection device 800 to complete all or part of the steps in the above-mentioned abnormal data detection method. The memory 802 is used to store various types of data to support operations on the anomaly data detection device 800. These data may include, for example, instructions for any application program or method operating on the anomaly data detection device 800, as well as application programs. Related data, such as contact data, messages sent and received, pictures, audio, video, etc. The memory 802 can be implemented by any type of volatile or non-volatile storage device or their combination, such as static random access memory (Static Random Access Memory, SRAM for short), electrically erasable programmable read-only memory ( Electrically Erasable Programmable Read-Only Memory (EEPROM for short), Erasable Programmable Read-Only Memory (EPROM for short), Programmable Read-Only Memory (PROM for short), read-only Memory (Read-Only Memory, ROM for short), magnetic memory, flash memory, magnetic disk or optical disk. Multimedia components 803 may include screen and audio components. The screen may be a touch screen, for example, and the audio component is used to output and/or input audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signal may be further stored in memory 802 or sent via communication component 805 . The audio component also includes at least one speaker for outputting audio signals. The I/O interface 804 provides an interface between the processor 801 and other interface modules. The other interface modules may be keyboards, mice, buttons, etc. These buttons can be virtual buttons or physical buttons. The communication component 805 is used for wired or wireless communication between the abnormal data detection device 800 and other devices. Wireless communication, such as Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G or 4G, or one or a combination of them, so the corresponding communication component 805 can include: Wi -Fi module, Bluetooth module, NFC module.

In an exemplary embodiment, the abnormal data detection device 800 may be configured by one or more application specific integrated circuits (Application Specific Integrated Circuits, ASICs for short), digital signal processors (Digital Signal Processors, DSPs for short), digital signal processing devices ( Digital Signal Processing Device (DSPD for short), Programmable Logic Device (PLD for short), Field Programmable Gate Array (FPGA for short), controller, microcontroller, microprocessor or other electronic components Implementation is used to perform one of the above-mentioned anomaly data detection methods.

In another exemplary embodiment, a computer-readable storage medium including program instructions is also provided. When the program instructions are executed by a processor, the steps of the above-mentioned abnormal data detection method are implemented. For example, the computer-readable storage medium can be the above-mentioned memory 802 including program instructions, and the above-mentioned program instructions can be executed by the processor 801 of the abnormal data detection device 800 to complete the above-mentioned abnormal data detection method.

Example 4

Corresponding to the above method embodiments, embodiments of the present disclosure also provide a readable storage medium. The readable storage medium described below and the abnormal data detection method described above may be mutually referenced.

A readable storage medium. A computer program is stored on the readable storage medium. When the computer program is executed by a processor, the steps of the abnormal data detection method of the above method embodiment are implemented.

The readable storage medium can specifically be a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (Random Access Memory, RAM), a magnetic disk or an optical disk that can store program codes. readable storage media.

The above descriptions are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and changes. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention. should be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims (8)

1. An abnormal data detection method, characterized by including: Obtain first information, which is merchandise sales data information of at least one store; Send the first information to the data preprocessing model to obtain second information, where the second information is price characteristic data information and sales volume characteristic data information obtained by preprocessing the first information; Send the second information to an anomaly detection model for abnormal data detection to obtain third information, where the third information is abnormal product sales data obtained by performing two clustering screenings on the second information; Send the third information to the verification module for processing to obtain fourth information, where the fourth information is the abnormal product sales data screened by the model after verifying the parameters; Wherein, the second information is sent to the anomaly detection model for abnormal data detection, and the third information is obtained, including: Send the price characteristic data information in the second information to the first clustering module for clustering to obtain the first clustering information, where the first clustering information is the abnormal data information in the price characteristic data information; Perform data correspondence mapping between the first clustering information and the sales volume characteristic data information in the second information to obtain second sub-information, where the second sub-information includes the sales volume characteristic data information corresponding to the first clustering information; Send the second sub-information to the second clustering module for processing to obtain second clustering information. The second clustering information is the abnormal product sales data obtained by performing two clustering screenings on the second information. ; Wherein, the second sub-information is sent to the second clustering module for processing to obtain the second clustering information, including: Data processing is performed based on the preset second initial parameters and the data information in the second sub-information, where the data information in the second sub-information is converted into coordinate data points in the spatial coordinate system, based on each According to the coordinate data points, the mutual reachable distance between each of the coordinate data points is obtained; Generate a weighted distance map based on the mutual reachability distance, and generate a minimum spanning tree of the mutual reachability distance based on the weighted distance map; Convert the minimum spanning tree into components of a hierarchical cluster structure according to the mutual reachability distance, and build a hierarchical cluster structure based on the components of the clustering hierarchy; The hierarchical cluster structure is compressed, and the data information in the second sub-information is classified based on the compressed hierarchical cluster structure to obtain abnormal data in the second sub-information. 2. The abnormal data detection method according to claim 1, characterized in that, the first information is sent to a data preprocessing model to obtain the second information, and the second information is obtained after preprocessing the first information. information, including: Perform data processing on the commodity sales data information, clear invalid data in the first information, and average-fill the incomplete data in the first information to obtain the first sub-information; Based on the first sub-information and the preset calculation formula, calculate the characteristic data of the first information, where the characteristic data of the first information includes price characteristic data and sales volume characteristic data; The feature data of the first information is normalized, and the normalized feature data is smoothed to obtain the preprocessed first information. 3. The abnormal data detection method according to claim 1, characterized in that the price characteristic data information in the second information is sent to the first clustering module for clustering to obtain the first clustering information, including: Traverse the price feature data information based on the preset first initial parameter information, and process the price feature data according to the cluster feature tree generation method in the BIRCH algorithm to obtain a price cluster feature tree; Obtain at least one clustering feature cluster based on the price clustering feature tree, and calculate the threshold range corresponding to each clustering feature cluster; Analyze all the threshold ranges, and use the minimum threshold range among all the threshold ranges as the normal threshold range for judging the normal point; Determine abnormal points in the price clustering feature tree based on the normal threshold range, and determine abnormal data information in the price feature data information based on the abnormal points. 4. An abnormal data detection device, characterized in that it includes: A first acquisition unit, configured to acquire first information, which is merchandise sales data information of at least one store; A first processing unit, configured to send the first information to a data preprocessing model to obtain second information, where the second information is price characteristic data information and sales volume characteristic data information obtained by preprocessing the first information; The second processing unit is used to send the second information to the anomaly detection model for abnormal data detection, and obtain the third information. The third information is the abnormal products obtained by performing two clustering screenings on the second information. sales data; A third processing unit, configured to send the third information to the verification module for processing to obtain fourth information, where the fourth information is the abnormal product sales data screened by the model after verifying the parameters; Wherein, the second processing unit includes: The first clustering subunit is used to send the price characteristic data information in the second information to the first clustering module for clustering to obtain the first clustering information, where the first clustering information is the price characteristic data. Abnormal data information in the information; The fourth processing subunit is used to perform data corresponding mapping between the first cluster information and the sales volume characteristic data information in the second information to obtain second sub-information, where the second sub-information includes the first cluster Sales volume characteristic data information corresponding to the information; The second clustering subunit is used to send the second sub-information to the second clustering module for processing to obtain the second clustering information. The second clustering information is performed twice on the second information. Abnormal product sales data obtained through clustering screening; Wherein, the second clustering subunit includes: The seventh clustering subunit is used to perform data processing based on the preset second initial parameters and the data information in the second sub-information, wherein the data information in the second sub-information is converted into a spatial coordinate system coordinate data points, and based on each of the coordinate data points, obtain the mutual reachable distance between each of the coordinate data points; An eighth clustering subunit is configured to generate a weighted distance map based on the mutual reachability distance, and generate a minimum spanning tree of the mutual reachability distance based on the weighted distance map; The ninth clustering subunit is used to convert the minimum spanning tree into components of a hierarchical cluster structure according to the mutual reachability distance, and build a hierarchical cluster structure based on the components of the clustering hierarchy; The tenth clustering subunit is used to compress the hierarchical cluster structure and classify the data information in the second sub-information based on the compressed hierarchical cluster structure to obtain abnormal data in the second sub-information. 5. The abnormal data detection device according to claim 4, characterized in that the device includes: The first processing subunit is used to perform data processing on the commodity sales data information, clear invalid data in the first information, and average-fill the incomplete data in the first information to obtain the first sub-unit. information; The second processing subunit is configured to calculate the characteristic data of the first information based on the first sub-information and the preset calculation formula. The characteristic data of the first information includes price characteristic data and sales volume characteristic data; The third processing subunit is used to normalize the characteristic data of the first information and smooth the normalized characteristic data to obtain the preprocessed first information. 6. The abnormal data detection device according to claim 4, characterized in that the device includes: The third clustering subunit is used to traverse the price feature data information based on the preset first initial parameter information, and process the price feature data according to the cluster feature tree generation method in the BIRCH algorithm to obtain price clustering feature tree; The fourth clustering subunit is used to obtain at least one clustering feature cluster based on the price clustering feature tree, and obtain and calculate the threshold range corresponding to each clustering feature cluster; The fifth clustering subunit is used to analyze all the threshold ranges, and use the minimum threshold range among all the threshold ranges as the normal threshold range for judging normal points; A sixth clustering subunit is configured to determine abnormal points in the price cluster feature tree based on the normal threshold range, and determine abnormal data information in the price feature data information based on the abnormal points. 7. An abnormal data detection device, characterized by including: Memory, used to store computer programs; A processor, configured to implement the steps of the abnormal data detection method according to any one of claims 1 to 3 when executing the computer program. 8. A readable storage medium, characterized in that: a computer program is stored on the readable storage medium, and when the computer program is executed by a processor, the abnormal data detection method according to any one of claims 1 to 3 is implemented. A step of.