CN115687958A

CN115687958A - Sample object determination method and device, electronic equipment and storage medium

Info

Publication number: CN115687958A
Application number: CN202211410100.2A
Authority: CN
Inventors: 熊建胜; 董莹莹; 李坤树; 李红云; 孙洋洋; 蒋迅婕
Original assignee: China United Network Communications Group Co Ltd
Current assignee: China United Network Communications Group Co Ltd
Priority date: 2022-11-11
Filing date: 2022-11-11
Publication date: 2023-02-03

Abstract

The application provides a sample object determination method, a sample object determination device, electronic equipment and a storage medium. The method comprises the following steps: acquiring a characteristic data set; mapping the characteristic attributes of the sample object in a preset multidimensional coordinate system to obtain a multidimensional coordinate point of the sample object in the multidimensional coordinate system; according to the multi-dimensional coordinate points, respectively carrying out first clustering processing and second clustering processing on the sample objects under different numbers of reference points in a multi-dimensional coordinate system to obtain a first clustering result and a second clustering result; determining a target clustering result according to the first clustering result and the second clustering result; determining a sample object to be detected belonging to the target classification in the sample objects and sample data to be detected according to the target clustering result; and detecting the sample data to be detected, and determining abnormal sample data and a target sample object. The method improves the efficiency of finding out the communication station with abnormal electricity charge.

Description

Sample object determination method and device, electronic equipment and storage medium

Technical Field

The present application relates to the field of machine learning technologies, and in particular, to a method and an apparatus for determining a sample object, an electronic device, and a storage medium.

Background

The communication station is used for storing a server for the communication enterprise, thereby providing IT services (IT services) for users and employees.

At present, when a communication station is in operation and use, electric charges are generated, the electric charges of the communication station can be generally acquired through a meter reading degree mode, but when the electric charges are acquired, the communication station may have the problem that part of electric meters are abnormal or the meter reading degree is wrong, and after the problem occurs, all the communication station stations need to be screened.

However, when screening is performed, because the number of the communication bureau stations is huge, the problem of difficulty in screening may exist, and therefore, the communication bureau stations with abnormal electricity charges cannot be found out quickly and accurately, the cost control of the whole enterprise is affected, and the requirement of the enterprise for maintaining normal operation cannot be met.

Disclosure of Invention

The application provides a sample object determining method, a sample object determining device, electronic equipment and a storage medium, which are used for solving the problems that a communication station with abnormal electricity charges cannot be found out quickly and accurately, the cost control of a whole enterprise is influenced, and the requirement of the enterprise for maintaining normal operation cannot be met.

In a first aspect, the present application provides a sample object determination method, including:

acquiring a characteristic data set, wherein the characteristic data set comprises a plurality of sample objects, characteristic attributes of the sample objects and sample data of the sample objects, and the characteristic attributes comprise a plurality of pieces of characteristic information;

mapping the characteristic attributes of the sample object in a preset multidimensional coordinate system to obtain multidimensional coordinate points of the sample object in the multidimensional coordinate system;

according to the multi-dimensional coordinate points, respectively carrying out first clustering processing and second clustering processing on the sample objects under different numbers of reference points in a multi-dimensional coordinate system to obtain a first clustering result and a second clustering result;

determining a target clustering result according to the first clustering result and the second clustering result;

determining sample objects to be detected belonging to the target classification in the sample objects and sample data to be detected of the sample objects to be detected according to the target clustering result;

and detecting the sample data to be detected, and determining abnormal sample data in the sample data to be detected and a target sample object corresponding to the abnormal sample data.

In an embodiment of the present application, acquiring a feature data set includes:

acquiring an initial characteristic data set, wherein the initial characteristic data set comprises a plurality of initial sample objects, initial characteristic attributes of the initial sample objects and initial sample data of the initial sample objects, and the initial characteristic attributes comprise a plurality of pieces of initial characteristic information;

and cleaning the initial characteristic data set to obtain a characteristic data set.

In this embodiment of the present application, the cleaning processing on the initial feature data set to obtain the feature data set includes:

determining abnormal initial characteristic information in the initial characteristic attribute;

determining an abnormal sample object and a sample object to be adjusted in the initial sample object according to the abnormal initial characteristic information;

deleting abnormal sample objects in the initial characteristic data set to obtain an initial adjustment characteristic data set;

and adjusting the sample object to be adjusted in the initial adjustment characteristic data set to obtain a characteristic data set.

In an embodiment of the present application, the initial feature data set comprises aligned sample objects;

adjusting the sample object to be adjusted in the initial adjustment feature data set to obtain a feature data set, including:

determining the information type of abnormal characteristic information in the sample object to be adjusted, wherein the abnormal characteristic information is the characteristic information in an abnormal state in the characteristic information;

determining comparison characteristic information in the comparison sample object according to the information type, wherein the comparison characteristic information is characteristic information corresponding to the information type of the abnormal characteristic information;

and adjusting the sample object to be adjusted in the initial adjustment characteristic data set according to the comparison characteristic information to obtain a characteristic data set.

In this embodiment of the present application, when the multidimensional coordinate system includes n first reference points, where n is a positive integer greater than 1, performing a first clustering process on the sample object to obtain a first clustering result, where the first clustering result includes:

determining a first sample object set of the first reference point according to the first reference point and the multi-dimensional coordinate point, wherein the first sample object set comprises a first sample object, and the first sample object is a sample object which meets a preset requirement in a multi-dimensional coordinate system and has a coordinate distance from the first reference point;

determining a first centroid point of the first sample object in the multi-dimensional coordinate system according to the first sample object set;

and repeating the step of determining the first sample object set of the first reference point according to the position relation between the first centroid point and the first reference point and the multidimensional coordinate point until the first centroid point and the first reference point meet the position relation, and obtaining a first clustering result.

In this embodiment, when the multidimensional coordinate system includes m second reference points, where m is a positive integer not equal to n, performing a second clustering process on the sample objects to obtain a second clustering result, includes:

determining a second sample object set of the second reference point according to the second reference point and the multi-dimensional coordinate point, wherein the second sample object set comprises a second sample object, and the second sample object is a sample object in the multi-dimensional coordinate system, and the coordinate distance between the second sample object and the second reference point meets the preset requirement;

determining a second centroid point of the second sample object in the multi-dimensional coordinate system according to the second sample object set;

and repeating the step of determining a second sample object set of the second reference point according to the second reference point and the multi-dimensional coordinate point according to the position relationship between the second centroid point and the second reference point until the second centroid point and the second reference point satisfy the position relationship, and obtaining a second clustering result.

In this embodiment of the present application, determining a target clustering result according to a first clustering result and a second clustering result includes:

acquiring a first contour coefficient according to the first clustering result;

acquiring a second contour coefficient according to the second clustering result;

comparing the first contour coefficient with the second contour coefficient:

when the first contour coefficient is larger than the second contour coefficient, determining that the first clustering result is a target clustering result;

and when the first contour coefficient is smaller than the second contour coefficient, determining the second clustering result as a target clustering result.

In the embodiment of the present application, the detecting processing of sample data to be detected, and determining abnormal sample data in the sample data to be detected and a target sample object corresponding to the abnormal sample data includes:

determining the defibering degree of the sample data to be detected;

determining abnormal sample data in the sample data to be detected according to the separation degree;

and determining a target sample object corresponding to the abnormal sample data according to the abnormal sample data.

In a second aspect, the present application provides a sample object determination apparatus, comprising:

the system comprises an acquisition module, a storage module and a processing module, wherein the acquisition module is used for acquiring a characteristic data set, and the characteristic data set comprises a plurality of sample objects, characteristic attributes of the sample objects and sample data of the sample objects, and the characteristic attributes comprise a plurality of pieces of characteristic information;

the obtaining module is used for mapping the characteristic attribute of the sample object in a preset multi-dimensional coordinate system to obtain a multi-dimensional coordinate point of the sample object in the multi-dimensional coordinate system;

the clustering module is used for respectively carrying out first clustering processing and second clustering processing on the sample objects under different numbers of reference points in a multi-dimensional coordinate system according to the multi-dimensional coordinate points to obtain a first clustering result and a second clustering result;

the first determining module is used for determining a target clustering result according to the first clustering result and the second clustering result;

the second determining module is used for determining a sample object to be detected in the sample objects, belonging to the target classification, and sample data of the sample object to be detected according to the target clustering result;

and the third determining module is used for detecting and processing the sample data of the sample object to be detected, and determining abnormal sample data in the sample data of the sample object to be detected and a target sample object corresponding to the abnormal sample data.

In a third aspect, the present application provides an electronic device, comprising: a processor, and a memory communicatively coupled to the processor;

the memory stores computer execution instructions;

the processor executes computer-executable instructions stored in the memory to implement the sample object determination method of the embodiments of the present application.

In a fourth aspect, a computer-readable storage medium has stored therein computer-executable instructions, which when executed by a processor, are used to implement the sample object determination method of the embodiments of the present application.

According to the sample object determining method, the sample object determining device, the electronic equipment and the storage medium, the characteristic data set is obtained, wherein the characteristic data set comprises a plurality of sample objects, characteristic attributes of the sample objects and sample data of the sample objects, and the characteristic attributes comprise a plurality of pieces of characteristic information; mapping the characteristic attributes of the sample object in a preset multidimensional coordinate system to obtain a multidimensional coordinate point of the sample object in the multidimensional coordinate system; according to the multi-dimensional coordinate points, respectively carrying out first clustering processing and second clustering processing on the sample objects under different numbers of reference points in a multi-dimensional coordinate system to obtain a first clustering result and a second clustering result; determining a target clustering result according to the first clustering result and the second clustering result; determining sample objects to be detected belonging to the target classification in the sample objects and sample data to be detected of the sample objects to be detected according to the target clustering result; the method comprises the steps of carrying out detection processing on sample data to be detected, determining abnormal sample data in the sample data to be detected and a target sample object corresponding to the abnormal sample data, so that the sample data of the sample objects belonging to a target classification can be obtained after clustering processing is carried out on the sample objects according to characteristic attributes, the sample data used for representing the electricity charge is closer due to higher similarity of the characteristic attributes, the sample data of the sample objects to be detected can be detected, the abnormal sample data can be obtained, the target sample object can be rapidly determined, so that workers can check the sample objects, meanwhile, in order to further improve the accuracy, the sample objects can be subjected to multiple different clustering processing, the sample objects to be detected with better clustering effect can be obtained, the accuracy of detecting the abnormal sample data is improved, a communication station with abnormal electricity charge can be rapidly and accurately found, the cost control of the whole communication enterprise is improved, and the effect of maintaining the normal operation of the communication enterprise is met.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the application and, together with the description, serve to explain the principles of the application.

Fig. 1 is a scene schematic diagram of a sample object determination method according to an embodiment of the present application;

fig. 2 is a schematic flowchart of a sample object determination method according to an embodiment of the present application;

fig. 3 is a schematic flowchart of another sample determination method provided in an embodiment of the present application;

fig. 4 is a schematic structural diagram of a sample determination device provided in an embodiment of the present application;

fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

With the above figures, there are shown specific embodiments of the present application, which will be described in more detail below. The drawings and written description are not intended to limit the scope of the inventive concepts in any manner, but rather to illustrate the concepts of the application by those skilled in the art with reference to specific embodiments.

Detailed Description

Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. The following description refers to the accompanying drawings in which the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the application, as detailed in the appended claims.

In the prior art, when detecting the abnormal electricity charge of the communication station, the detection is generally performed by setting a rule, for example, a fixed threshold is set artificially, and when the monthly electricity charge of the communication station is more than 30% higher than the monthly electricity charge, the abnormal electricity charge of the communication station is considered. However, when the communication station is detected in this way, because the number of the communication station stations is large, and the usage conditions of each communication station are different, for example, the usage amount of electricity in summer and in winter is suddenly increased, and the increase or decrease of the devices in the communication station causes the significant change of the electricity consumption, the communication station with a problem cannot be accurately determined only by setting a rule, so that the cycle of abnormal detection of the communication station is further long, the cost control of the whole communication enterprise cannot be met, and the effect of maintaining the normal operation requirement of the communication enterprise cannot be met.

According to the method and the device, based on the use condition of the communication station, the communication station is clustered by taking objective factors such as the equipment condition and the use environment of the communication station as the characteristic attributes of the communication station, and the communication station is classified into the same communication station, abnormal sample data and a target sample object are quickly and accurately found by comparing the sample data, so that a worker can further check the communication station corresponding to the target sample object according to the abnormal sample data and the target sample object, and whether the communication station has a problem is determined.

The embodiment of the application provides a sample object determination method and device, electronic equipment and a storage medium.

Fig. 1 is a scene schematic diagram of a sample object determination method according to an embodiment of the present application. As shown in fig. 1, an execution subject of the sample object determination method may be a server. The server can be a mobile phone, a tablet, a computer and other devices. The implementation manner of the execution subject is not particularly limited in this embodiment, as long as the execution subject can obtain the feature data set, and map the feature attribute of the sample object in the preset multidimensional coordinate system to obtain the multidimensional coordinate point of the sample object in the multidimensional coordinate system; respectively carrying out first clustering processing and second clustering processing on the sample objects under different numbers of reference points in a multi-dimensional coordinate system according to the multi-dimensional coordinate points to obtain a first clustering result and a second clustering result; determining a target clustering result according to the first clustering result and the second clustering result; determining sample objects to be detected belonging to the target classification in the sample objects and sample data to be detected of the sample objects to be detected according to the target clustering result; and detecting the sample data to be detected, and determining the abnormal sample data in the sample data to be detected and the target sample object corresponding to the abnormal sample data.

The scrubbing process may be Data scrubbing (Data scrubbing), which may refer to a process of finding and correcting recognizable errors in a Data file, including checking Data consistency, processing invalid and missing values, and the like.

Clustering may refer to the process of dividing a collection of sample objects into classes composed of similar objects.

Fig. 2 is a schematic flowchart of a sample object determination method according to an embodiment of the present application. The execution subject of the method may be a server or other servers, and this embodiment is not limited herein, and as shown in fig. 2, the method may include:

s201, acquiring a characteristic data set, wherein the characteristic data set comprises a plurality of sample objects, characteristic attributes of the sample objects and sample data of the sample objects, and the characteristic attributes comprise a plurality of characteristic information.

The sample object may refer to an object that needs to be subjected to screening determination, where the object may be a person or an object, and in this embodiment, the object may be a communication station.

The characteristic attribute may refer to attribute information of the sample object, and the characteristic attribute is composed of a plurality of characteristic information, for example, in some embodiments, when the power consumption of the user needs to be determined, the characteristic attribute may include power consumption time and power consumption duration, and the power consumption time and the power consumption duration may be the characteristic information. In this embodiment, when the sample object is a communication station, the characteristic attribute may include information such as the number and power of servers of the communication station, the number and power of switches, and the number and power of air conditioners.

The sample data may refer to data that needs to be detected in the sample object, for example, in some embodiments, the sample data may be power usage of the user. In this embodiment, the sample data may be an electric charge of the communication station.

In some embodiments, the number of sample data is at least one.

The feature data set may be composed of a plurality of sample objects, and a feature attribute and sample data of each sample object. In this embodiment, the number of the communication bureau stations may be multiple, that is, the number of the sample objects may correspond to multiple.

In the embodiment of the application, the feature data set can be obtained through a plurality of sample objects, feature attributes and sample data. The method for acquiring the characteristic data set can comprise the step of obtaining the characteristic data set by receiving a sample object input by a user and characteristic attributes and sample data of the sample object.

In some embodiments, the characteristic attribute of the sample object in the feature data set may be a characteristic attribute whose similarity satisfies a similarity requirement, for example, when the environment of the a region and the B region is greatly different, which results in that the characteristic attribute of the sample object in the a region is greatly different from the characteristic attribute of the sample object in the B region, the sample object in the feature data set may include only the sample object in the a region or only the sample object in the B region.

In an embodiment of the present application, a method for acquiring a feature data set may include:

Wherein the initial feature data set may be a plurality of initial sample objects, and an initial feature attribute of each initial sample object and a set of initial sample data of the initial sample object.

The initial sample objects may refer to unprocessed sample objects received, and the number of initial sample objects may be greater than or equal to the number of sample objects. In some embodiments, unprocessed may refer to a sample object that has not undergone a cleaning process on the initial characteristic attributes of the initial sample object and the sample data.

Sample data may refer to data in the original sample object that needs to be detected.

The initial sample data may refer to data that needs to be detected in the initial sample object.

In this embodiment of the present application, the cleaning processing on the initial feature data set may refer to a processing procedure of removing an initial sample object with obviously abnormal initial feature attributes and an initial sample object with too much missing initial feature attributes in the initial feature data set, so as to obtain the feature data set.

In this embodiment, the method for obtaining the feature data set by performing a cleaning process on the initial feature data set may include:

For example, in some embodiments, when the initial feature information in the initial feature attribute includes the number of air conditioners, when the number of air conditioners is 3.5, the initial feature information is abnormal initial feature information with abnormal data, and when the number of air conditioners is not filled, the initial feature information is abnormal initial feature information with missing data.

In the embodiment of the present application, the initial characteristic information may be determined by a preset rule, for example, the number of the air conditioners is a positive integer and cannot exceed 20. By the determination of the rule, the sample object determination means can automatically determine whether the initial characteristic information is abnormal initial characteristic information.

The abnormal sample object may be an initial sample object with abnormal initial characteristic information, and in this embodiment of the present application, the abnormal sample object may be an initial sample object with at least one piece of characteristic information with data abnormality, or an initial sample object with existing characteristic information with data missing more than a preset threshold.

Since there is much abnormal feature information or missing feature information in the abnormal sample object, the abnormal sample object usually has no means for adjustment and correction, and the abnormal sample object in the initial feature data set is deleted to obtain an initial adjusted feature data set. The deleting process refers to a process of deleting the abnormal sample object, the initial characteristic attribute of the abnormal sample object and the initial sample data.

The sample object to be adjusted may be an adjustable initial sample object, and in this embodiment of the present application, the sample object to be adjusted may be an initial sample object whose existing characteristic information of data loss is smaller than a preset threshold.

The sample object to be adjusted only has a small amount of characteristic information missing, so the sample object to be adjusted can be subjected to adjustment processing. The adjustment processing refers to a process of adjusting missing feature information in the sample object to be adjusted according to the initial feature attributes of other sample objects in the initial adjustment feature data set.

In this embodiment, the initial feature data set includes comparison sample objects, and the comparison sample objects may refer to initial sample objects with no missing or abnormal feature information. The method for adjusting the sample object to be adjusted in the initial adjustment feature data set to obtain the feature data set may include:

The information type may refer to an attribute type of the abnormal feature information, for example, when the abnormal feature information is the number of air conditioners, the information type may be the number of air conditioners.

The comparison characteristic information may be characteristic information corresponding to the attribute type of the abnormal characteristic information in the comparison sample object, for example, in some embodiments, when the information type of the abnormal characteristic information is the number of air conditioners, the comparison characteristic information may be the number of air conditioners.

The adjustment processing refers to assigning values to the abnormal feature information in the sample object to be adjusted according to all the comparison feature information, for example, in some embodiments, when the information type of the abnormal feature information may be the number of air conditioners, and the comparison feature information is the number of air conditioners, the abnormal feature information may be assigned by obtaining a median value of the number of air conditioners in all the other comparison feature information, so as to obtain the adjusted sample object to be adjusted.

And after the sample object to be adjusted is obtained, taking the data of the sample object to be adjusted and the data of the comparison sample object as a characteristic data set.

S202, mapping the characteristic attributes of the sample object in a preset multidimensional coordinate system to obtain multidimensional coordinate points of the sample object in the multidimensional coordinate system.

The mapping of the characteristic attribute of each sample object in the preset multidimensional coordinate system may refer to mapping the characteristic attribute of the sample object in the multidimensional coordinate system with the number of dimensions corresponding to the number of types according to the number of types of the characteristic information in the characteristic attribute. For example, in some embodiments, when the sample object feature attribute types include the number of servers, power, the number of switches, power, the number of air conditioners, and power, since there are six data types, the feature attribute may be mapped in a six-dimensional coordinate system, so as to obtain a multi-dimensional coordinate point corresponding to each sample object.

S203, according to the multi-dimensional coordinate points, respectively carrying out first clustering processing and second clustering processing on the sample objects under different numbers of reference points in the multi-dimensional coordinate system to obtain a first clustering result and a second clustering result.

The clustering process is a process of dividing a set of sample objects into a plurality of classes composed of similar objects. For example, in the embodiment of the present application, when the sample object is a communication station, and the characteristic attribute is information such as the number and power of servers, the number and power of switches, the number and power of air conditioners, and the like, performing clustering on the sample object refers to a process of classifying the sample object according to the similarity of the characteristic attribute. The clustering process may be a clustering process of the reference points performed on the sample object according to positions of the reference points in the multidimensional coordinate system, where the clustering process of the reference points may refer to determining multidimensional coordinate points close to the reference points according to the positions of the reference points, and obtaining a first clustering result and a second clustering result for classifying the sample object according to distribution of the multidimensional coordinate points.

The clustering result may refer to one or more similar object components obtained after clustering, for example, in the embodiment of the present application, when there are two reference points, two results of sample objects belonging to different categories may be obtained after clustering the sample objects; when there are three reference points, after clustering the sample objects, three results of sample objects belonging to different categories can be obtained.

In this embodiment of the present application, when the multidimensional coordinate system includes n first reference points, where n is a positive integer greater than 1, the method for performing the first clustering process on the sample object to obtain the first clustering result may include:

determining a first sample object set of the first reference point according to the first reference point and the multi-dimensional coordinate point, wherein the first sample object set comprises a first sample object, and the first sample object is a sample object which meets a preset requirement in a multi-dimensional coordinate system and has a coordinate distance with the first reference point;

The number and the position of the first reference points can be reference points randomly selected in a multi-dimensional coordinate system.

The first sample object may be obtained by determining whether the distance between the multidimensional coordinate point and the first reference point meets a preset requirement according to the position of the first reference point in the multidimensional coordinate system after the first reference point is obtained. The distance between the multidimensional coordinate point and the first reference point can be obtained through the Euclidean distance, and the preset requirement can be a preset threshold requirement or a preset requirement determined by comparing the distance between each multidimensional coordinate point and each first reference point.

In an embodiment of the present application, each first reference point includes a first sample object set corresponding thereto, that is, each first reference point includes a plurality of first sample objects corresponding thereto.

After determining the plurality of first sample objects, first centroid points of the plurality of first sample objects may be determined from the plurality of first sample objects, and the first centroid points may be points at which distances to the corresponding respective first sample objects are shortest.

After the first centroid point is obtained, comparing the position relationship between the first centroid point and the first reference point, wherein in some embodiments, the position relationship may include intersection and non-intersection, and when the position relationship is non-intersection, the current first centroid point may be used as the first reference point to perform the step of determining the first sample object set of the first reference point according to the first reference point and the multidimensional coordinate point, so that until the first centroid point and the first reference point satisfy the position relationship.

And when the first centroid point and the first reference point meet the position relation, representing that the first clustering result of the first clustering process is the optimal result of the first clustering process. Wherein the first clustering result includes n classified sample objects.

In this embodiment of the present application, when the multidimensional coordinate system includes m second reference points, where m is a positive integer not equal to n, the second clustering process is performed on the sample object, and the method for obtaining the second clustering result may include:

The number and the positions of the second datum points can be randomly selected datum points, and the number of the second datum points is different from the number of the first datum points.

The second sample object may be obtained by determining whether a distance between the multidimensional coordinate point and the second reference point meets a preset requirement according to a position of the second reference point in the multidimensional coordinate system after the second reference point is obtained. The distance between the multidimensional coordinate point and the second reference point can be obtained through the Euclidean distance, and the preset requirement can be a preset threshold requirement or a distance determined by comparing each multidimensional coordinate point with each second reference point.

In an embodiment of the present application, each of the second fiducial points includes a second set of sample objects corresponding thereto, i.e., each of the second fiducial points includes a plurality of second sample objects corresponding thereto.

After determining the plurality of second sample objects, second centroid points of the plurality of second sample objects may be determined from the plurality of second sample objects, and the second centroid points may be points at which distances to the corresponding respective second sample objects are shortest.

After the second centroid point is obtained, the position relationship between the second centroid point and the second reference point may be compared, where in some embodiments, the position relationship may include intersection and disjointness, and when the position relationship is disjointness, the step of determining the second sample object set of the second reference point according to the second reference point and the multidimensional coordinate point may be performed using the current second centroid point as the second reference point, so that until the second centroid point and the second reference point satisfy the position relationship.

And when the second centroid point and the second reference point meet the position relation, representing that the second clustering result of the second clustering processing is the optimal result of the second clustering processing. Wherein the second clustering result includes m classified sample objects.

And S204, determining a target clustering result according to the first clustering result and the second clustering result.

And the target clustering result is the clustering result with better clustering effect in the first clustering result and the second clustering result. The clustering effect can be determined by contour Coefficient (Silhouette Coefficient), CH Score (Calinski Harabasz Score), davison baudin index (DBI, davies bouldin Score), and the like.

In this embodiment of the present application, when determining a clustering effect through a contour coefficient, the method for determining a target clustering result according to a first clustering result and a second clustering result may include:

acquiring a first contour coefficient according to the first clustering result;

comparing the first contour coefficient with the second contour coefficient:

In the embodiment of the present application, in order to obtain the most accurate clustering result, multiple different clustering processes may be performed on the sample object, where the number of the set reference points is different in each clustering process, for example, the number of the reference points may be 2 to 20, when the clustering processes are performed, the clustering results with the number of the reference points being 2 to 20 may be respectively obtained, then the contour coefficients with the number of the reference points being 2 to 20 may be respectively obtained, and finally, the target clustering result with the largest contour coefficient is selected.

In the embodiment of the application, the target clustering result can be obtained by pairwise comparison.

S205, determining sample objects to be tested belonging to the target classification in the sample objects and sample data to be tested of the sample objects to be tested according to the target clustering result;

s206, detecting the sample data to be detected, and determining abnormal sample data in the sample data to be detected and a target sample object corresponding to the abnormal sample data.

The target clustering result can be characterized as sample objects belonging to a plurality of classifications, wherein each classification can be screened respectively when screening is carried out according to the target clustering result, namely, the target classification is selected from the plurality of classifications in sequence.

According to the target clustering result and the sample data of the sample object, the data that the sample data in the sample object of the same classification is abnormal can be intuitively determined, and the target sample object can be determined according to the data. For example, in the embodiment of the present application, when it is determined that the number, power, number, power of switches, number, and power of air conditioners belong to the same classification of communication stations, the electricity charge data of each communication station under the classification may be obtained, and since the characteristic attributes of each communication station under the classification are similar, the electricity charge data should also be within a certain range, and abnormal detection is performed on the electricity charges of each communication station under the classification, so as to determine an abnormal communication station that may have a problem, thereby facilitating a worker to check the abnormal communication station.

The anomaly detection processing can be detection through an isolated forest algorithm model, namely, the data to be detected of each sample object to be detected can be input into the isolated forest algorithm model, so that the target data in the data to be detected can be rapidly determined, and the target sample object in the sample object to be detected can be determined according to the target data in the data to be detected.

For example, in the embodiment of the present application, the method for performing detection processing on sample data to be detected, and determining abnormal sample data in the sample data to be detected and a target sample object corresponding to the abnormal sample data may include:

determining the defibering degree of the sample data to be detected;

The sample determining method provided by the embodiment of the application can obtain the sample objects to be detected belonging to the target classification in the sample objects after clustering the sample objects according to the characteristic attributes, and can perform abnormal detection on the sample data of the sample objects to be detected to obtain abnormal sample data because the characteristic attributes have higher similarity and the sample data used for representing the electricity charge are more similar, so that the target sample objects can be quickly determined to facilitate the detection of workers.

Fig. 3 is a schematic flowchart of another sample determination method provided in an embodiment of the present application, and as shown in fig. 3, the method includes:

s301, acquiring resource data and monthly electric charge data of the communication station, wherein the resource data comprises the number and power of devices in the communication station.

The devices in the communication station may include a server, an exchange, an air conditioner, and the like.

S302, cleaning the resource data and the monthly electric charge data to obtain the cleaned resource data and the cleaned monthly electric charge data.

The cleaning processing may refer to a processing procedure of removing data with obvious abnormality and excessive missing values, acquiring median of other similar data, and filling a communication station with a small number of missing values.

S303, constructing a cluster model of the cleaned resource data.

The clustering model may be a model using a K-Means clustering algorithm (K-Means clustering algorithm), the input of the clustering model may be resource data, and the output may be a classification result for the communication station.

S304, determining the number of the classifications of the communication station stations and the communication station stations under each classification according to the clustering model.

The number of the classifications may be the number of the clusters in a K-Means clustering algorithm, and the method for determining the number of the classifications may include: presetting the number of the classifications to be 2-20, and clustering each classification number to obtain a clustering result; acquiring a contour coefficient corresponding to each clustering result; selecting a target contour coefficient with the maximum contour coefficient; and determining the number of the classifications corresponding to the target contour coefficient as the optimal number of the classifications, and determining the communication bureau station under each classification according to the optimal number of the classifications.

The optimal classification number can represent the closest equipment power and quantity of the communication station under the same classification.

S305, acquiring monthly electric charge data of the communication station stations in the same classification, and inputting the monthly electric charge data of the communication station stations in the same classification into an abnormality detection model to obtain abnormal monthly electric charge data of each classification and the communication station stations in an abnormal state.

The communication station under the same classification can be determined to be in an abnormal state by inputting the monthly electric charge data of the communication station under the same classification into the abnormal detection model.

The abnormal detection model can be a model adopting an isolated forest algorithm, and abnormal monthly electric charge data can be rapidly determined from a large amount of monthly electric charge data through the isolated forest algorithm. Thereby causing the sample object determining means to determine the communication office station in an abnormal state based on the abnormal monthly electricity rate data.

And S306, sending the information of the communication station in the abnormal state so that the staff can check the monthly electricity charge data of the communication station in the abnormal state.

After determining the information of the communication bureau station in the abnormal state, the information of the communication bureau station in the abnormal state can be sent to the terminal, so that a worker can check monthly electricity charge data of the communication bureau station in the abnormal state, the communication bureau station in the abnormal state can be corrected and adjusted, the communication bureau station in the abnormal state can be screened quickly and accurately, the cost control of the whole communication enterprise is influenced, and the requirement of the communication enterprise for maintaining normal operation cannot be met.

Fig. 4 is a schematic structural diagram of a sample determination device according to an embodiment of the present application. As shown in fig. 4, the sample determining apparatus 40 for the interest point includes: an obtaining module 401, an obtaining module 402, a clustering module 403, a first determining module 404, a second determining module 405, and a third determining module 406. Wherein:

the obtaining module 401 is configured to obtain a feature data set, where the feature data set includes a plurality of sample objects, a plurality of feature attributes of the sample objects, and sample data of the sample objects, and the feature attributes include a plurality of pieces of feature information;

an obtaining module 402, configured to map a characteristic attribute of the sample object in a preset multidimensional coordinate system, so as to obtain a multidimensional coordinate point of the sample object in the multidimensional coordinate system;

a clustering processing module 403, configured to perform first clustering processing and second clustering processing on the sample objects under different numbers of reference points in the multi-dimensional coordinate system according to the multi-dimensional coordinate points, respectively, to obtain a first clustering result and a second clustering result;

a first determining module 404, configured to determine a target clustering result according to the first clustering result and the second clustering result;

a second determining module 405, configured to determine, according to the target clustering result, a sample object to be detected belonging to the target classification in the sample objects, and sample data of the sample object to be detected;

the third determining module 406 is configured to perform detection processing on the sample data of the sample object to be detected, and determine abnormal sample data in the sample data of the sample object to be detected and a target sample object corresponding to the abnormal sample data.

In this embodiment of the application, the obtaining module 401 may further be configured to:

In this embodiment, the obtaining module 401 may be further configured to:

In this embodiment of the application, the clustering module 403 may be further configured to:

In this embodiment, the cluster processing module 403 may be further configured to:

In this embodiment, the first determining module 404 may be further configured to:

acquiring a first contour coefficient according to the first clustering result;

comparing the first contour coefficient with the second contour coefficient:

In this embodiment of the application, the third determining module 406 may specifically be configured to:

determining the defibering degree of the sample data to be detected;

As can be seen from the above, the sample determining apparatus of this embodiment includes an obtaining module 401, configured to obtain a feature data set, where the feature data set includes a plurality of sample objects, a plurality of feature attributes of the sample objects, and sample data of the sample objects, and the feature attributes include a plurality of pieces of feature information; an obtaining module 402, configured to map the characteristic attribute of the sample object in a preset multidimensional coordinate system, to obtain a multidimensional coordinate point of the sample object in the multidimensional coordinate system; the clustering module 403 is configured to perform first clustering processing and second clustering processing on the sample objects under different numbers of reference points in the multidimensional coordinate system according to the multidimensional coordinate points, and the first determining module 404 is configured to determine a target clustering result according to the first clustering result and the second clustering result; the second determining module 405 is configured to determine, according to the target clustering result, a sample object to be detected belonging to the target classification in the sample objects, and sample data of the sample object to be detected; the third determining module 406 is configured to perform detection processing on the sample data of the sample object to be detected, and determine abnormal sample data in the sample data of the sample object to be detected and a target sample object corresponding to the abnormal sample data. Therefore, after the sample objects are clustered according to the characteristic attributes, the sample objects to be detected belonging to the target classification in the sample objects can be obtained, and due to the fact that the similarity of the characteristic attributes is high, sample data used for representing the electric charge is close to each other, abnormal detection can be conducted between the sample data of the sample objects to be detected, abnormal sample data can be obtained, the target sample objects can be determined quickly, and accordingly workers can check the sample objects.

Fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present application. As shown in fig. 5, the electronic apparatus 50 includes:

the electronic device 50 may include components such as a processor 501 of one or more processing cores, memory 502 of one or more computer-readable storage media, a communications component 503, and so on. The processor 501, the memory 502, and the communication unit 503 are connected by a bus 504.

In particular implementations, the at least one processor 501 executes computer-executable instructions stored by the memory 502 to cause the at least one processor 501 to perform the sample object determination method as described above.

For a specific implementation process of the processor 501, reference may be made to the above method embodiments, which implement the similar principle and technical effect, and this embodiment is not described herein again.

In the embodiment shown in fig. 5, it should be understood that the Processor may be a Central Processing Unit (CPU), other general purpose processors, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), etc. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of a method disclosed in connection with the present invention may be embodied directly in a hardware processor, or in a combination of the hardware and software modules within the processor.

The Memory may include a high-speed Memory (RAM) and may also include a Non-volatile Memory (NVM), such as at least one disk Memory.

The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended ISA (EISA) bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, the buses in the figures of the present application are not limited to only one bus or one type of bus.

In some embodiments, a computer program product is also proposed, which comprises a computer program or instructions, which when executed by a processor, implement the steps of any of the above-mentioned sample object determination methods.

The above operations can be implemented in the foregoing embodiments, and are not described in detail herein.

It will be understood by those skilled in the art that all or part of the steps of the methods of the above embodiments may be performed by instructions, or by instructions controlling associated hardware, which may be stored in a computer-readable storage medium and loaded and executed by a processor.

To this end, embodiments of the present application provide a computer-readable storage medium, in which a plurality of instructions are stored, where the instructions can be loaded by a processor to execute the steps in any one of the sample object determination methods provided in the embodiments of the present application.

Wherein the storage medium may include: a Read Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and the like.

According to an aspect of the application, a computer program product or computer program is provided, comprising computer instructions, the computer instructions being stored in a computer readable storage medium.

Since the instructions stored in the storage medium may execute the steps in any sample object determination method provided in the embodiments of the present application, beneficial effects that can be achieved by any sample object determination method provided in the embodiments of the present application may be achieved, which are detailed in the foregoing embodiments and will not be described again here.

Other embodiments of the present application will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the application and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the application being indicated by the following claims.

It will be understood that the present application is not limited to the precise arrangements that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the application is limited only by the appended claims.

Claims

1. A method for sample object determination, the method comprising:

mapping the characteristic attribute of the sample object in a preset multidimensional coordinate system to obtain a multidimensional coordinate point of the sample object in the multidimensional coordinate system;

according to the multi-dimensional coordinate points, respectively carrying out first clustering processing and second clustering processing on the sample objects under different numbers of reference points in the multi-dimensional coordinate system to obtain a first clustering result and a second clustering result;

determining a sample object to be detected belonging to the target classification in the sample objects and sample data to be detected of the sample object to be detected according to the target clustering result;

2. The method of claim 1, wherein the obtaining the feature data set comprises:

and cleaning the initial characteristic data set to obtain the characteristic data set.

3. The method of claim 2, wherein performing a cleaning process on the initial feature data set to obtain the feature data set comprises:

determining abnormal initial characteristic information in the initial characteristic attributes;

deleting the abnormal sample object in the initial characteristic data set to obtain an initial adjustment characteristic data set;

and adjusting the sample object to be adjusted in the initial adjustment characteristic data set to obtain the characteristic data set.

4. The method of claim 3, wherein the initial set of feature data comprises aligned sample objects;

the adjusting the sample object to be adjusted in the initial adjustment feature data set to obtain the feature data set includes:

and adjusting the sample object to be adjusted in the initial adjustment characteristic data set according to the comparison characteristic information to obtain the characteristic data set.

5. The method of claim 1, wherein when the multi-dimensional coordinate system includes n first reference points, where n is a positive integer greater than 1, performing a first clustering process on the sample objects to obtain a first clustering result, comprises:

determining a first sample object set of the first reference point according to the first reference point and the multi-dimensional coordinate point, wherein the first sample object set comprises a first sample object, and the first sample object is a sample object in the multi-dimensional coordinate system, and the coordinate distance between the sample object and the first reference point meets a preset requirement;

and repeating the step of determining a first sample object set of the first reference point according to the first reference point and the multi-dimensional coordinate point according to the position relationship between the first centroid point and the first reference point until the first centroid point and the first reference point meet the position relationship to obtain a first clustering result.

6. The method of claim 5, wherein when the multi-dimensional coordinate system includes m second reference points, m being a positive integer not equal to n, performing a second clustering process on the sample object to obtain a second clustering result, comprising:

determining a second sample object set of the second reference point according to the second reference point and the multi-dimensional coordinate point, wherein the second sample object set comprises a second sample object, and the second sample object is a sample object in the multi-dimensional coordinate system, and the coordinate distance between the second sample object and the second reference point meets a preset requirement;

and according to the position relation between the second centroid point and the second reference point, repeating the step of determining a second sample object set of the second reference point according to the second reference point and the multi-dimensional coordinate point until the second centroid point and the second reference point meet the position relation, and obtaining a second clustering result.

7. The method of claim 1, wherein determining a target clustering result from the first clustering result and the second clustering result comprises:

acquiring a first contour coefficient according to the first clustering result;

comparing the first contour coefficient and the second contour coefficient:

and when the first contour coefficient is smaller than the second contour coefficient, determining that the second clustering result is a target clustering result.

8. The method according to claim 1, wherein the performing the detection processing on the sample data to be detected, determining the abnormal sample data in the sample data to be detected, and the target sample object corresponding to the abnormal sample data comprises:

determining the defibering degree of the sample data to be detected;

according to the separation degree, determining abnormal sample data in the sample data to be detected;

9. A sample object determination apparatus, comprising:

the obtaining module is used for mapping the characteristic attribute of the sample object in a preset multidimensional coordinate system to obtain a multidimensional coordinate point of the sample object in the multidimensional coordinate system;

the clustering processing module is used for respectively carrying out first clustering processing and second clustering processing on the sample object under different numbers of reference points in the multi-dimensional coordinate system according to the multi-dimensional coordinate points to obtain a first clustering result and a second clustering result;

the second determining module is used for determining a sample object to be detected in the sample objects, belonging to the target classification, and sample data to be detected of the sample object to be detected according to the target clustering result;

and the third determining module is used for detecting the sample data to be detected, determining abnormal sample data in the sample data to be detected and a target sample object corresponding to the abnormal sample data.

10. An electronic device, comprising: a processor, and a memory communicatively coupled to the processor;

the memory stores computer-executable instructions;

the processor executes computer-executable instructions stored by the memory to implement the sample object determination method of any of claims 1 to 8.

11. A computer-readable storage medium having computer-executable instructions stored thereon, which when executed by a processor, are configured to implement the sample object determination method according to any one of claims 1 to 8.