CN116401566A - Grouping method, grouping device, computer equipment and storage medium for data objects - Google Patents

Abstract

The invention relates to the field of data processing, and discloses a grouping method, a grouping device, computer equipment and a storage medium of data objects, wherein the grouping method comprises the following steps: dividing the candidate data object into a plurality of data object groups by adopting a spatial clustering method based on geographic coordinate data; performing data object elimination on any two data object groups based on the inter-group distance to obtain a plurality of data object groups; sampling the multiple groups of data object groups to obtain a comparison data object group and an experimental data object group; performing experiments of the control data object group and the experimental data object group, and calculating target variable difference values of the control data object group and the experimental data object group after the experiments; when the difference value of the target variable is larger than or equal to the difference value threshold, exchanging partial data objects in the control data object group and the experimental data object group, and eliminating the mutual influence of the data of the experimental object and the data of the experimental object which are not involved in the invention, so that the control group and the experimental group with larger individual difference have comparability.

Description

Data object grouping method, device, computer equipment and storage medium

technical field

The present invention relates to the technical field of data processing, in particular to a data object grouping method, device, computer equipment and storage medium.

Background technique

Controlled experiments are to prove whether the experimental treatment factors produce certain results (the Change or difference) research methods are widely used in medicine, biology, agronomy, engineering science, Internet technology and even socio-economic fields. The key process of carrying out controlled experiments is to ensure the comparability of the "experimental group" and "control group". Usually, random sampling is used when grouping. The samples of the two groups are independent of each other and the individual differences are small.

However, in many fields of research, it is usually difficult for researchers to control too much on the subjects participating in the experiment, and there is inevitably a spatial correlation between the sample data, that is, individual data with close geographical locations will affect each other, while There may be large characteristic differences between individual data at a long distance (for example, in real estate price research, we want to compare the impact of different pricing schemes on the final income, the prices of different houses in the same real estate will affect each other, and the prices of different houses in the same real estate will affect each other, while Houses in different real estate projects may not be comparable in price due to large differences in location, transportation, environment, supporting facilities, and apartment types.) This spatial correlation and individual variability caused great difficulty in grouping data from controlled experiments.

Therefore, it is urgent to provide a control experiment data grouping scheme that can eliminate the mutual influence of the data of subjects participating in the experiment and the data of subjects not participating in the experiment, so that the "control group" and "experimental group" with large individual differences are comparable .

Contents of the invention

In view of this, the present invention provides a data object grouping method, device, computer equipment, and storage medium to solve the problem that it is difficult to control the objects participating in the experiment too much in many fields of research, so that the sample data There are inevitably spatial correlations and individual differences among them, which cause extremely difficult problems for the grouping of controlled experimental data objects.

In a first aspect, the present invention provides a method for grouping data objects, the method comprising: based on geographic coordinate data, using a spatial clustering method to divide candidate data objects into multiple data object groups; Any two data object groups in a data object group are eliminated to obtain multiple groups of data object groups that are not correlated in space between groups; random sampling is performed on multiple groups of data object groups to obtain the control data object group and the experimental data object group group; the control data object group and the experimental data object group are processed in a controlled experiment, and the target variable difference value between the control data object group and the experimental data object group at the current observation point is calculated after the experiment; when the target variable difference value is greater than the difference value When the threshold is equal to or equal to the difference threshold, some data objects in the control data object group and the experimental data object group are exchanged. Through the above process, the mutual influence between the data of subjects participating in the experiment and the data of subjects not participating in the experiment can be eliminated, so that the "control group" and "experimental group" with large individual differences are comparable.

In a second aspect, the present invention provides a device for grouping data objects according to an embodiment of the present invention. The evaluation device mainly includes: a data object division module, a data object elimination module, a data object sampling module, a data object experiment module, and the data object exchange module, wherein the data object division module is used to divide the candidate data objects into multiple data object groups based on the geographic coordinate data by adopting the spatial clustering method; Any two data object groups in multiple data object groups are eliminated to obtain multiple groups of data object groups that are not correlated in space between groups; the data object sampling module is used to perform random sampling processing on multiple groups of data object groups to obtain comparison The data object group and the experimental data object group; the data object experiment module is used to process the control data object group and the experimental data object group in a controlled experiment, and calculate the target of the control data object group and the experimental data object group at the current observation point after the experiment The variable difference value; data object exchange module is used to exchange some data objects in the control data object group and the experimental data object group when the target variable difference value is greater than or equal to the difference value threshold. Through the above process, the mutual influence between the data of subjects participating in the experiment and the data of subjects not participating in the experiment can be eliminated, so that the "control group" and "experimental group" with large individual differences are comparable.

In a third aspect, the present invention provides a computer device, including: a memory and a processor, the memory and the processor are connected to each other, computer instructions are stored in the memory, and the processor performs the first aspect by executing the computer instructions or the method for grouping data objects in any of the corresponding implementations.

In a fourth aspect, the present invention provides a computer-readable storage medium, where computer instructions are stored on the computer-readable storage medium, and the computer instructions are used to cause a computer to execute the data object in the above-mentioned first aspect or any corresponding implementation manner grouping method.

Description of drawings

In order to more clearly illustrate the specific implementation of the present invention or the technical solutions in the prior art, the following will briefly introduce the accompanying drawings that need to be used in the specific implementation or description of the prior art. Obviously, the accompanying drawings in the following description The drawings show some implementations of the present invention, and those skilled in the art can obtain other drawings based on these drawings without any creative effort.

FIG. 1 is a schematic diagram of an application environment according to an embodiment of the present invention;

2 is a schematic flow diagram of a method for grouping data objects according to an embodiment of the present invention;

Fig. 3 is a schematic flow chart of another method for grouping data objects according to an embodiment of the present invention;

Fig. 4 is a structural block diagram of a device for grouping data objects according to an embodiment of the present invention;

FIG. 5 is a schematic diagram of a hardware structure of a computer device according to an 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 It is a part of embodiments of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making creative efforts belong to the protection scope of the present invention.

Please refer to Fig. 1, Fig. 1 is a schematic diagram of an application environment provided by the embodiment of the present application, the schematic diagram includes a client 10 and a server 20, wherein the server 10 receives the geographic coordinate data and the alternative data objects uploaded by the client 20 Finally, the candidate data objects can be grouped based on the geographic coordinate data, and finally the control experimental data that eliminates the spatial correlation and makes the "control group" and "experimental group" with large individual differences comparable grouping scheme.

Specifically, in the embodiment of the present application, the client 10 shown in FIG. 1 that transmits geographic coordinate data and alternative data objects may be a user's smart phone, desktop computer, tablet computer, notebook computer, digital assistant, smart wearable device and other types of physical devices; among them, smart wearable devices may include smart bracelets, smart watches, smart glasses, smart helmets, etc. Of course, the client 10 is not limited to the above-mentioned electronic device with a certain entity, and it can also be software running in the above-mentioned electronic device. For example, the client 10 can be a web page or an application provided by the service provider to the user.

Optionally, the client 10 may include a display screen, a storage device, and a processor connected through a data bus. Wherein, the display screen is used for geographic coordinate data and alternative data objects, and the display screen may be a touch screen of a mobile phone or a tablet computer, etc. The storage device is used to store geographical coordinate data and alternative data objects or other data materials, etc., and the storage device can be the internal memory of the client 10, or can be a smart media card (smart media card), a secure digital card (securedigital card) , flash memory card (flash card) and other storage devices. Processors can be single-core or multi-core processors.

In the embodiment of the present application, the server 20 shown in FIG. 1 may receive the geographical coordinate data and the candidate data objects, or it may also be other computer terminals having the same functions as the server, or similar computing devices. Further, the server 20 can be replaced with a server system, a computing platform or a server cluster including multiple servers.

According to an embodiment of the present invention, an embodiment of a method for grouping data objects is provided. It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and, Although a logical order is shown in the flowcharts, in some cases the steps shown or described may be performed in an order different from that shown or described herein.

In this embodiment, a method for grouping data objects is provided, which can be used for the above-mentioned clients, such as smart phones, tablet computers, etc. FIG. 2 is a flowchart of a method for grouping data objects according to an embodiment of the present invention, as shown in FIG. 2, the process includes the following steps:

Step S201, based on the geographic coordinate data, the candidate data objects are divided into multiple data object groups by using a spatial clustering method.

In this embodiment, firstly, the candidate data objects are obtained according to the experimental objectives, and then based on the geographical coordinate data, the candidate data objects are divided into multiple data object groups by using the spatial clustering method, so as to provide a data basis for the subsequent division of the data object groups. Wherein, the candidate data objects include the spatial position of each object and corresponding object information. That is, the candidate data objects are data objects that are spatial and can reflect the attributes of the experimental target.

In an optional implementation, when there are m candidate data objects, 2n data objects are randomly selected from the m candidate data objects (the value of n needs to be determined according to the specific problem to be studied and the candidate data The number of objects m is determined), as the initial cluster center (that is, to determine the initial coordinates of the cluster center). For example, 2×200=400 data objects can be randomly selected from m=10000 candidate data objects (where 200 is the number of circles to which m=10000 candidate data objects belong), as the initial clustering center. Calculate the distance between the remaining candidate data objects and each initial cluster center, and divide each candidate data object into its nearest class, thus realizing the division of candidate data objects into pairs of data object groups. Wherein, in the process of calculating the distance between each candidate data object and the initial cluster center, it can be calculated according to the geographic coordinate data of each candidate data object, the geographic coordinate data in the initial cluster, and the radius of the earth.

Step S202, based on the inter-group distance, perform data object elimination on any two data object groups in the plurality of data object groups, to obtain multiple groups of data object groups that are spatially irrelevant between the groups.

In this embodiment, after obtaining multiple data object groups, in order to prevent the spatial correlation between each data object group, it is necessary to perform data analysis on any two data object groups in the multiple data object groups based on the inter-group distance. Object culling, resulting in groups of data object groups that are spatially uncorrelated between groups. Wherein, when performing data object elimination, groups with a large number of data objects are preferentially eliminated.

In an optional implementation, for the obtained data object groups, calculate their minimum distance two by two, and eliminate the distance between groups that is less than a preset distance threshold (the distance threshold is related to whether there is spatial correlation, less than the distance The threshold has spatial correlation, otherwise it does not have spatial correlation and needs to be set according to the research question). That is, the above-mentioned elimination process is performed in pairs of all data object groups until the minimum distance between any two experimental groups is not less than the distance threshold, that is, multiple data object groups that are spatially uncorrelated between groups are obtained.

For example, when the preset distance threshold is 1500m. If among the obtained 400 data object groups, two groups a and b contain 120 data objects and 150 data objects respectively, and there are two data objects in group a and group b: data object 1 and data table object 2 If the distance is the shortest (assuming 1200 meters), data object 2 in group b needs to be deleted (because group b contains more data objects). Calculate the distance of the data objects in the two groups a and b in this way and delete the data objects until the shortest distance between the data objects in the two groups is not less than 1500 meters.

In step S203, random sampling is performed on multiple groups of data subject groups to obtain a control data subject group and an experimental data subject group.

In this embodiment, when multiple groups of data object groups are obtained, there is no spatial correlation between the multiple groups of data object groups at this time. Group Data subject groups are divided into control data subject groups and experimental data subject groups.

In step S204, the control data object group and the experimental data object group are subjected to a control experiment process, and the difference value of the target variable between the control data object group and the experimental data object group at the current observation point is calculated after the experiment.

In this embodiment, when multiple groups of data object groups are randomly sampled to obtain the control data object group and the experimental data object group, due to the existence of individual differences, the control data object group and the experimental data object group obtained at this time There is a high probability that the target variable of . Therefore, it is necessary to process the control data object group and the experimental data object group in a controlled experiment, and calculate the target variable difference between the control data object group and the experimental data object group at the current observation point after the experiment. Among them, the target variable is the value of the experimental target after the experiment, that is, the next observation point.

In an optional embodiment, 2n data object groups are randomly divided into 2 groups (each group contains n data object groups, such as can be divided into 2 groups, each group contains n=200 data objects), respectively as a control Data Object Group and Experiment Data Object Group. Due to the existence of individual differences, the target variables of the control data subject group and the experimental data subject group obtained at this time are likely to be different. Without loss of generality, it is assumed that at the observation time point after the experimental treatment, the target variable for judging the difference between the control data subject group and the experimental data subject group is the sum of the individual target variables in the group (it can also be the mean, median, etc. quantity). That is to say, we need to compare the sum of the individual target variables in the group when the control data object group and the experimental data object group are at the current observation point, and compare the sum of the individual target variables in the group when the control data object group is at the current observation point with the experimental data object group at The difference between the sum of the individual target variables in the group at the current observation point is obtained to obtain the target variable difference between the control data object group and the experimental data object group at the current observation point.

Step S205, when the difference value of the target variable is greater than or equal to the difference threshold, some data objects in the control data object group and the experimental data object group are exchanged.

In this embodiment, in order to reduce the difference between the control data object group and the experimental data object group in the control experiment, the relationship between the target variable difference value and the difference value threshold is judged after the control experiment, and when the target variable difference value is greater than the difference When the value threshold is equal to or equal to the difference value threshold, some data objects in the control data object group and the experimental data object group are exchanged. Wherein, in the process of exchanging some data objects in the control data object group and the experimental data object group, it may be an object in the control data object group and the experimental data object group, or it may be an exchange of multiple objects. Change.

The method for grouping data objects provided in this embodiment divides candidate data objects into multiple data object groups by using spatial clustering based on geographical coordinate data, so as to provide a data basis for the division of subsequent data object groups; Distance, remove data objects from any two data object groups in multiple data object groups to obtain multiple groups of data object groups that are not spatially correlated between groups; perform random sampling processing on multiple groups of data object groups to obtain comparison The data object group and the experimental data object group; the control data object group and the experimental data object group are processed in the control experiment, and the difference value of the target variable between the control data object group and the experimental data object group at the current observation point is calculated after the experiment. When the target When the variable difference value is greater than the difference value threshold or equal to the difference value threshold, some data objects in the control data object group and the experimental data object group are exchanged to reduce the difference between the control data object group and the experimental data object group in the control experiment In order to eliminate the interaction between the data of subjects participating in the experiment and the data of subjects not participating in the experiment, the "control group" and "experimental group" with large individual differences are comparable.

In this embodiment, a method for grouping data objects is provided, which can be used for the above-mentioned mobile terminals, such as smart phones, tablet computers, etc. FIG. 3 is a flowchart of a method for grouping data objects according to an embodiment of the present invention, as shown in FIG. 3, the process includes the following steps:

Step S301, based on the geographic coordinate data, the candidate data objects are divided into multiple data object groups by using a spatial clustering method.

Specifically, the above step S301 includes:

In step S3011, a plurality of objects are randomly selected from candidate data objects as initial clustering centers.

In an optional implementation, when there are m candidate data objects, 2n data objects are randomly selected from the m candidate data objects (the value of n needs to be determined according to the specific problem to be studied and the candidate data The number of objects m is determined), as the initial cluster center (that is, to determine the initial coordinates of the cluster center). For example, 2×200=400 data objects can be randomly selected from m=10000 candidate data objects (where 200 is the number of circles to which m=10000 candidate data objects belong), as the initial clustering center.

Step S3012, calculate the distance from the remaining candidate data objects to each initial cluster center, and divide the candidate data objects into the classes corresponding to the initial cluster centers.

In an optional implementation, first obtain the coordinate latitude and longitude of each data object in the candidate data object, the coordinate longitude and latitude of each initial cluster center, and the radius of the earth; then calculate the function based on the spherical distance, and each of the candidate data objects The coordinate longitude and latitude of the data object, the coordinate longitude and latitude of each initial clustering center, and the radius of the earth, create a clustering distance calculation model; finally, calculate the distance from the remaining candidate data objects to each initial clustering center through the clustering distance calculation model .

The expression of the clustering distance calculation model is:

d＝distCosin(lng _i ,lat _i ,lng ^O ,lat ^O ,r)

Among them, distCosin is the spherical distance calculation function, lng _i is the coordinate longitude of the candidate data object i, lat _i is the coordinate latitude of the candidate data object i, lng ^o is the cluster center coordinate longitude of the oth class, and lat ^o is The latitude of the cluster center coordinates of the oth class, r is the radius of the earth, and the default is 6378137 meters.

Step S3013, update the cluster center of each class based on the division result to form multiple data object groups.

In an optional implementation manner, the longitude data of each candidate data object in the current class is obtained, and the longitude data of each candidate data object are summed and divided by the number of candidate data objects contained in the current class to obtain the current The longitude of the cluster center coordinates of the class; obtain the latitude data of each candidate data object in the current class, and divide the latitude data of each candidate data object by the number of candidate data objects contained in the current class to obtain the current class The latitude and longitude of the cluster center coordinates of the current class; update the cluster center of the current class according to the longitude and latitude of the cluster center coordinates of the current class. Step S3012 to step S3013 are repeated until the cluster center of each class does not change, and a plurality of data object groups are obtained.

The expression of the cluster center of the current class is:

Among them, lng ^o is the longitude of the clustering center coordinates of the oth class; lat ^o is the latitude and longitude of the clustering center coordinates of the oth class; N _o is the number of data objects contained in the oth class; lng _i is the number of data objects belonging to the oth class The coordinate longitude of the data object i of the first class; lat _i is the coordinate latitude of the data object i belonging to the oth class.

Step S302, based on the inter-group distance, perform data object elimination on any two data object groups in the plurality of data object groups, to obtain multiple groups of data object groups that are spatially irrelevant between the groups.

Specifically, first obtain the coordinate longitude and latitude of each data object in the first comparison data object group, the coordinate longitude and latitude of each data object in the second comparison data object group, and the radius of the earth; then calculate the function based on the spherical distance, and the first comparison data object The coordinate longitude and latitude of each data object in the group, the coordinate longitude and latitude of each data object in the second comparison data object group, and the radius of the earth, create a distance calculation model between groups; through the distance calculation model between groups, calculate the first comparison data object group and the second comparison data object group The distance between the two closest data objects in the two-comparison data object group. When the distance between the two closest data objects in the first comparison data object group and the second comparison data object group is less than the distance threshold, delete the first comparison data object group and the second comparison data object group. One of two data objects, resulting in multiple groups of data object subgroups that are spatially uncorrelated between groups.

In an optional implementation, for the obtained data object groups, calculate the minimum distance two by two, and eliminate the distance between the groups that is less than the preset distance threshold D (the distance threshold is related to whether there is spatial correlation, less than the The distance threshold has spatial correlation, otherwise it does not have spatial correlation and needs to be set according to the research question). For example, for two data object groups o ₁ and o ₂ , the minimum distance is the distance of the closest data object between the two groups:

Among them, distCosin is the spherical distance calculation function, lng _i and lat _i are the coordinate longitude and latitude of data object i in group o ₁ respectively, lng _j and lat _j are the coordinate longitude and latitude of data object j in group o ₂ respectively , r is the radius of the earth, the default is 6378137 meters.

小于预先设置的距离阈值D(该距离阈值与是否具有空间相关性有关，小于该距离阈值则具有空间相关性，否则不具有空间相关性，需根据研究的问题设定)，则将o₁和o₂两个数据对象组之间距离最近的2个数据对象删除1个(优先删除所属分组中包含数据对象数量较多的那个)。即若if

is less than the preset distance threshold D (this distance threshold is related to whether there is spatial correlation, if it is smaller than this distance threshold, it has spatial correlation, otherwise it does not have spatial correlation, it needs to be set according to the research problem), then o ₁ and o ₂ Delete one of the two data objects with the closest distance between the two data object groups (priority to delete the one that contains more data objects in the group to which it belongs). That is if

and

Then remove j. Where N _o1 and N _o2 are respectively the number of data objects included in the _o1th and _o2th data object groups.

Perform the above elimination process in pairs for all data object groups, until the minimum distance between any two data object groups is not less than D, that is, 2n data object groups that are not spatially correlated between groups are obtained.

In step S303, random sampling is performed on multiple groups of data subject groups to obtain a control data subject group and an experimental data subject group.

For details, please refer to step S203 in the embodiment shown in FIG. 2 , which will not be repeated here.

In step S304, the control data object group and the experimental data object group are subjected to a control experiment process, and the target variable difference value between the control data object group and the experimental data object group at the current observation point is calculated after the experiment.

Specifically, the above step S304 includes:

Step S3041, obtaining the value of the target variable of the control data object group at the current observation point, and the value of the target variable of the experimental data object group at the current observation point.

In an optional implementation, it is assumed that at the observation time point s+1 after the experimental treatment, it is judged that the target variable of the control data subject group and the experimental data subject group is the sum of the individual target variables in the group (it can also be the mean value , median and other statistics). That is, we need to compare:

Y _A,s+1 =∑ _i∈A y _i,s+1

Y _B,s+1 ＝∑ _j∈B y _j,s+1

Among them, A is the experimental data object group; B is the control data object group; y _{i, s+1} (i∈A) represents the value of the target variable of the data object belonging to the experimental data object group at the observation time point s+1 ; y _{j, s+1} (j∈B) represents the value of the target variable of the data object belonging to the control data object group at the observation time point s+1; Y _{A, s+1} represents the target variable of the experimental data object group The value at the observation time point s+1; Y _{B, s+1} represents the value of the target variable of the control data object group at the observation time point s+1. Similarly, we can calculate the target variable values of the control data object group and the experimental data object group in the historical period t(t=1,2,...,s):

Y _A,t =∑ _i∈A y _i,t

Y _B,t =∑ _j∈B y _j,t

Among them, A is the experimental data object group; B is the control data object group; y _i,t (i∈A) represents the target variable of the data object belonging to the experimental data object group at the observation time point t (t=1,2,. ..,s) when the value is taken; y _j,t (j∈B) indicates that the target variable of the data object belonging to the control data object group is at the observation time point t(t=1,2,...,s) Y _A,s+1 represents the value of the target variable of the experimental data object group at the observation time point t(t=1,2,...,s); Y _B,s+1 represents the control data The value of the target variable of the object group at the observation time point t (t=1,2,...,s).

Step S3042, according to the difference between the value of the target variable of the control data object group at the current observation point and the value of the target variable of the experimental data object group at the current observation point, obtain the control data object group and the experimental data object group at the current observation point The target variable difference value for the point.

In an optional implementation, the maximum difference in the value of the target variable between the experimental data subject group and the control data subject group in each period is calculated:

If ΔY is greater than the preset threshold Δ (the threshold needs to be specifically set according to the research question), then extract a data subject group from the control data subject group and the experimental data subject group to exchange groups and recalculate ΔY until the If ΔY is less than the preset threshold Δ, the control data object group and the experimental data object group that can be used in the control experiment are obtained.

For example, a certain city has m=10000 houses for rent as experimental objects. The latitude and longitude coordinates (lng _i , lat _i ) and historical rental income y _i,t of each house are known. Our purpose is To select two groups of houses in the experimental data object group and the control data object group from this batch of houses, adopt two different pricing schemes A and B respectively, and observe the rental income of the two groups of houses under different pricing schemes y _{i,s +1} for the level of variance.

First, 2×200=400 suites can be randomly selected from m=10,000 suites (where 200 is the number of business districts to which 10,000 suites belong), as the initial clustering center. Calculate the distance from the remaining listings to each cluster center, and divide each listing into the class closest to it, and update the clustering center of each class based on the division results.

Secondly, a distance threshold D=1500 meters can be set. If in the obtained 400 listing groups, there are two groups a and b containing 120 and 150 listings respectively, and there are two listings in group a and group b—house 1 and house 2 are the closest (assuming it is 1200 meters), you need to delete listing 2 in group b (because group b contains more listings). Calculate the distance between the listings in groups a and b in this way and delete the listings until the shortest distance between the listings in the two groups is not less than D=1500 meters.

Finally, the difference value threshold Δ=50000 yuan can be set. If the experimental data object group and the control data object group are obtained, and the maximum difference in total rental income in each historical month exceeds 50,000 yuan, then extract from the experimental data object group and the control data object group (each containing n=200 groups of house sources) Group 1 houses are exchanged until the maximum difference in total rental income between the experimental data subject group and the control data subject group in each historical month does not exceed 50,000 yuan. At this time, adopt two pricing schemes A and B for the experimental data subject group and the control data subject group respectively, and then observe the total monthly rental income of the experimental data subject group and the control data subject group. If the difference is significantly higher than 50,000 yuan or higher than Three times the standard deviation of the difference between the experimental data object group and the control data object group in the total rental income of each month, it shows that there are significant differences between the two pricing schemes A and B, and the scheme with high total rental income is significantly better than the one with low total rental income plan.

Step S305, when the difference value of the target variable is greater than or equal to the difference threshold, some data objects in the control data object group and the experimental data object group are exchanged.

For details, please refer to step S205 in the embodiment shown in FIG. 2 , which will not be repeated here.

The method for grouping data objects provided in this embodiment divides candidate data objects into multiple data object groups by using spatial clustering based on geographical coordinate data, so as to provide a data basis for the division of subsequent data object groups; Distance, remove data objects from any two data object groups in multiple data object groups to obtain multiple groups of data object groups that are not spatially correlated between groups; perform random sampling processing on multiple groups of data object groups to obtain comparison The data object group and the experimental data object group; the control data object group and the experimental data object group are processed in the control experiment, and the difference value of the target variable between the control data object group and the experimental data object group at the current observation point is calculated after the experiment. When the target When the variable difference value is greater than the difference value threshold or equal to the difference value threshold, some data objects in the control data object group and the experimental data object group are exchanged to reduce the difference between the control data object group and the experimental data object group in the control experiment In order to eliminate the interaction between the data of subjects participating in the experiment and the data of subjects not participating in the experiment, and to control the difference of the historical data of the target variable between the control data subject group and the experimental data subject group, so that the "control group" with large individual differences group" and "experimental group" are comparable.

In this embodiment, a device for grouping data objects is also provided, and the device is used to implement the above embodiments and preferred implementation modes, and those that have already been described will not be repeated. As used below, the term "module" may be a combination of software and/or hardware that realizes a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementations in hardware, or a combination of software and hardware are also possible and contemplated.

This embodiment provides a device for grouping data objects, as shown in Figure 4, including:

The data object division module 401 is configured to divide the candidate data objects into multiple data object groups by adopting the spatial clustering method based on the geographic coordinate data.

In some optional implementation manners, the data object division module 401 includes:

The object extraction unit is used to randomly extract multiple data objects from candidate data objects as initial cluster centers.

The object division unit is used to calculate the distance from the remaining candidate data objects to each initial cluster center, and divide the candidate data objects into the classes corresponding to the initial cluster centers.

Specifically, first obtain the coordinate longitude and latitude of each data object in the candidate data object, the coordinate longitude and latitude of each initial cluster center, and the radius of the earth; then calculate the function based on the spherical distance, and the coordinate longitude and latitude of each data object in the candidate data object, The coordinate longitude and latitude of each initial cluster center, and the radius of the earth, create a cluster distance calculation model; finally, calculate the distance from the remaining candidate data objects to each initial cluster center through the cluster distance calculation model.

The cluster update unit is configured to update the cluster center of each class based on the division result to form multiple data object groups.

Specifically, first obtain the longitude data of each candidate data object in the current class, and divide the longitude data of each candidate data object by the number of candidate data objects contained in the current class to obtain the cluster center of the current class Coordinate longitude; then obtain the latitude data of each candidate data object in the current class, and divide the latitude data of each candidate data object by the number of candidate data objects contained in the current class to obtain the cluster center of the current class Coordinate latitude; finally, update the cluster center of the current class according to the coordinate longitude and latitude of the cluster center coordinates of the current class.

The data object elimination module 402 is configured to perform data object elimination on any two data object groups in the plurality of data object groups based on the inter-group distance, so as to obtain multiple groups of data object groups that are not spatially correlated between the groups.

In some optional implementation manners, the data object elimination module 402 includes:

The first data acquisition unit is used to acquire the coordinate longitude and latitude of each data object in the first comparison data object group, the coordinate longitude and latitude of each data object in the second comparison data object group, and the radius of the earth;

The second data acquisition unit is used to calculate the function based on the spherical distance, and the coordinate longitude and latitude of each data object in the first comparison data object group, the coordinate longitude and latitude of each data object in the second comparison data object group, and the radius of the earth to create a group space distance calculation model;

The distance calculation unit is used to calculate the distance between the two closest data objects in the first comparison data object group and the second comparison data object group through the inter-group distance calculation model;

The object deletion unit is used to delete the first comparison data object group and the second comparison data object group when the distance between the two closest data objects in the first comparison data object group and the second comparison data object group is less than the distance threshold. One of the two closest data objects in the object group.

The data object sampling module 403 is used to perform random sampling processing on multiple groups of data object groups to obtain a control data object group and an experimental data object group.

The data object experiment module 404 is used to perform the control experiment processing on the control data object group and the experimental data object group, and calculate the target variable difference value between the control data object group and the experimental data object group at the current observation point after the experiment.

In some optional embodiments, the data object experiment module 404 includes:

The observation point value acquisition unit obtains the value of the target variable of the control data object group at the current observation point, and the value of the target variable of the experimental data object group at the current observation point;

The observation point difference calculation unit is used to obtain the control data object group and the experimental The target variable difference value of the data object group at the current observation point.

The data object exchange module 405 is used for exchanging some data objects in the control data object group and the experimental data object group when the difference value of the target variable is greater than or equal to the difference value threshold, and then the photo of the object exchange The data subject group and the experimental data subject group conduct a controlled experiment.

The device for grouping data objects in this embodiment is presented in the form of a functional unit, where a unit refers to an ASIC circuit, a processor and a memory that execute one or more software or fixed programs, and/or other devices that can provide the above functions device.

The further functional descriptions of the above-mentioned modules and units are the same as those of the above-mentioned corresponding embodiments, and will not be repeated here.

An embodiment of the present invention also provides a computer device, which has the apparatus for grouping data objects shown in FIG. 4 above.

Please refer to FIG. 5. FIG. 5 is a schematic structural diagram of a computer device provided in an optional embodiment of the present invention. As shown in FIG. 5, the computer device includes: one or more processors 10, memory 20, and a Interfaces of various components, including high-speed interfaces and low-speed interfaces. The various components are communicatively connected to each other using different buses and may be mounted on a common motherboard or otherwise as desired. The processor may process instructions executed within the computer device, including instructions stored in or on the memory, to display graphical information of a GUI on an external input/output device such as a display device coupled to an interface. In some alternative implementations, multiple processors and/or multiple buses may be used with multiple memories and multiple memories, if desired. Likewise, multiple computing devices may be connected, with each device providing some of the necessary operations (eg, as a server array, a set of blade servers, or a multi-processor system). A processor 10 is taken as an example in FIG. 5 .

Processor 10 may be a central processing unit, a network processor or a combination thereof. Wherein, the processor 10 may further include a hardware chip. The aforementioned hardware chip may be an application specific integrated circuit, a programmable logic device or a combination thereof. The aforementioned programmable logic device may be complex programmable logic device, field programmable logic gate array, general array logic or any combination thereof.

Wherein, the memory 20 stores instructions executable by at least one processor 10, so that at least one processor 10 executes and realizes the methods shown in the above-mentioned embodiments.

The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and an application program required by at least one function; created data, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some optional implementations, the memory 20 may optionally include memories that are remotely located relative to the processor 10, and these remote memories may be connected to the computer device through a network. Examples of the aforementioned networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state disk; the memory 20 may also include a combination of the above types of memories.

The computer device also includes a communication interface 30 for the computer device to communicate with other devices or a communication network.

The embodiment of the present invention also provides a computer-readable storage medium. The above-mentioned method according to the embodiment of the present invention can be implemented in hardware or firmware, or can be recorded in a storage medium, or can be downloaded through the network for original storage. Computer code on a remote storage medium or a non-transitory machine-readable storage medium to be stored on a local storage medium so that the methods described herein can be stored on a computer using a general purpose computer, a special purpose processor, or programmable or dedicated hardware Such software processing on storage media. Wherein, the storage medium may be a magnetic disk, an optical disk, a read-only memory, a random access memory, a flash memory, a hard disk or a solid-state hard disk, etc.; further, the storage medium may also include a combination of the above types of memories. It can be understood that a computer, processor, microprocessor controller or programmable hardware includes a storage component that can store or receive software or computer code, and when the software or computer code is accessed and executed by the computer, processor or hardware, the above-mentioned implementation Example method.

Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the present invention. within the bounds of the requirements.

Claims (10)

1. A method of grouping data objects, the method comprising:

dividing the candidate data object into a plurality of data object groups by adopting a spatial clustering method based on geographic coordinate data;

based on the inter-group distance, eliminating the data object of any two data object groups in the plurality of data object groups to obtain a plurality of groups of data object groups with uncorrelated inter-group space;

randomly sampling the multiple groups of data object groups to obtain a comparison data object group and an experimental data object group;

performing control experiment treatment on the control data object group and the experimental data object group, and calculating target variable difference values of the control data object group and the experimental data object group at the current observation point after experiments;

and exchanging part of the data objects in the control data object group and the experimental data object group when the difference value of the target variable is larger than or equal to the difference value threshold.

2. The method of claim 1, wherein the step of dividing the candidate data objects into a plurality of data object groups using spatial clustering based on the geographic coordinate data comprises:

randomly extracting a plurality of objects from the candidate data objects to serve as initial clustering centers;

calculating the distance from the rest candidate data objects to each initial clustering center, and dividing the candidate data objects into classes corresponding to the initial clustering centers;

and updating the clustering center of each class based on the division result to form the plurality of data object groups.

3. The method according to claim 2, wherein the step of calculating the distance of the remaining candidate data objects to each initial cluster center comprises:

acquiring the coordinate longitude and latitude of each data object in the candidate data objects, the coordinate longitude and latitude of each initial clustering center and the earth radius;

creating a clustering distance calculation model based on a spherical distance calculation function, and the longitude and latitude of coordinates of each data object in the candidate data objects, the longitude and latitude of coordinates of each initial clustering center and the radius of the earth;

and calculating the distance from the rest candidate data objects to each initial clustering center through the clustering distance calculation model.

4. The method of claim 2, wherein the step of updating the cluster center of each class based on the division result comprises:

acquiring longitude data of each candidate data object in the current class, summing the longitude data of each candidate data object, and dividing the summed longitude data by the number of the candidate data objects contained in the current class to obtain cluster center coordinate longitudes of the current class;

acquiring latitude data of each candidate data object in the current class, summing the latitude data of each candidate data object, and dividing the sum by the number of the candidate data objects contained in the current class to obtain the cluster center coordinate latitude of the current class;

and updating the cluster center of the current class according to the cluster center coordinate longitude and the cluster center coordinate latitude of the current class.

5. The method of claim 1, wherein the step of performing data object culling on any two of the plurality of data object groups based on inter-group distance comprises:

acquiring the longitude and latitude of coordinates of each data object in the first comparison data object group, the longitude and latitude of coordinates of each data object in the second comparison data object group and the radius of the earth;

creating an inter-group distance calculation model based on a spherical distance calculation function, and the longitude and latitude of the coordinates of each data object in the first comparison data object group, the longitude and latitude of the coordinates of each data object in the second comparison data object group, and the earth radius;

Calculating the distance between two data objects closest to each other in the first comparison data object group and the second comparison data object group through the inter-group distance calculation model;

and deleting one of the two data objects closest to the first contrast data object group and the second contrast data object group when the distance between the two data objects closest to the first contrast data object group and the second contrast data object group is smaller than a distance threshold value.

6. The method of claim 1, wherein the step of calculating the target variable difference values of the control data object group and the experimental data object group at the current observation point after the experiment comprises:

obtaining the value of a target variable of the comparison data object group at the current observation point and the value of a target variable of the experimental data object group at the current observation point;

and obtaining the difference value of the target variables of the comparison data object group and the experimental data object group at the current observation point according to the difference value of the target variables of the comparison data object group and the experimental data object group at the current observation point.

7. The method of any one of claims 1 to 6, wherein after exchanging the extracted partial data objects from each of the set of illumination data objects and the set of experiment data objects, the method further comprises:

performing a control experiment on the control data object group and the experimental data object group.

8. A data object grouping apparatus, the apparatus comprising:

the data object dividing module is used for dividing the candidate data objects into a plurality of data object groups by adopting a spatial clustering method based on geographic coordinate data;

the data object eliminating module is used for eliminating data objects of any two data object groups in the plurality of data object groups based on the inter-group distance to obtain a plurality of groups of data object groups with uncorrelated inter-group space;

the data object sampling module is used for carrying out random sampling processing on the plurality of groups of data object groups to obtain a comparison data object group and an experimental data object group;

the data object experiment module is used for carrying out contrast experiment treatment on the contrast data object group and the experimental data object group, and calculating the target variable difference value of the contrast data object group and the experimental data object group at the current observation point after the experiment;

And the data object interchange module is used for interchanging partial data objects in the data object group and the experimental data object group when the difference value of the target variable is larger than or equal to the difference value threshold value.

9. A computer device, comprising:

a memory and a processor in communication with each other, the memory having stored therein computer instructions which, upon execution, cause the processor to perform the method of any of claims 1 to 7.

10. A computer readable storage medium having stored thereon computer instructions for causing a computer to perform the method of any one of claims 1 to 7.

Images