CN108595669A - A kind of unordered classified variable processing method and processing device - Google Patents
A kind of unordered classified variable processing method and processing device Download PDFInfo
- Publication number
- CN108595669A CN108595669A CN201810403750.1A CN201810403750A CN108595669A CN 108595669 A CN108595669 A CN 108595669A CN 201810403750 A CN201810403750 A CN 201810403750A CN 108595669 A CN108595669 A CN 108595669A
- Authority
- CN
- China
- Prior art keywords
- variable
- unordered
- classification
- classified variable
- unordered classified
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Image Analysis (AREA)
Abstract
This application provides a kind of unordered classified variable processing method and processing devices, wherein this method includes：Obtain unordered classified variable collection, wherein unordered classified variable collection includes the unordered classified variable of at least two classes, and corresponding dependent variable is two classified variables；For every a kind of unordered classified variable that unordered classified variable is concentrated, it is classification accounting of the unordered classified variable of target classification value in two classified variables in such unordered classified variable to count dependent variable value in such unordered classified variable；Based on the classification accounting of all kinds of unordered classified variables, clustering processing is carried out to unordered classified variable collection, obtains multiple unordered classified variable subsets；Wherein, each unordered classified variable subset includes at least a kind of unordered classified variable, and each unordered classified variable subset is a corresponding orderly classified variable.Grouping can be realized without the participation of artificial experience in the application so that the efficiency of packet transaction is higher, and further improves the objectivity and accuracy of group result.
Description
Technical field
This application involves computer processing technical field, in particular to a kind of unordered classified variable processing method and
Device.
Background technology
With the arrival in big data epoch and the rapid development of Internet technology, explosion type is presented in the data volume of every profession and trade
Increase.In these data, the classified variable of unordered type accounts for a big chunk proportion.For the ease of potential in mining data
Value, needs to handle above-mentioned unordered classified variable using effective processing method.Wherein, variable grouping problem is variable
Common problem in processing method.
It is grouped problem for variable, what is studied mostly is the grouping to continuous variable.For the classified variable of unordered type
The correlative study of grouping is less, can take two kinds of processing methods substantially：First, being grouped by experience, second is that without grouping
Directly bring use.
However, less efficient above by the method that experience is grouped, and cannot ensure the validity of grouping, it is above-mentioned not
The method directly used is grouped then for the relatively broad more classified variables of classifying, cannot ensure subsequently to model etc. and answer
The effect of used time.
Invention content
In view of this, the embodiment of the present application is designed to provide a kind of unordered classified variable processing method and processing device,
While raising is grouped the efficiency of processing to unordered classified variable, the objectivity and accuracy of group result are also improved.
In a first aspect, the embodiment of the present application provides a kind of unordered classified variable processing method, the method includes：
Obtaining unordered classified variable collection, wherein the unordered classified variable collection includes the unordered classified variable of at least two classes, and
Corresponding dependent variable is two classified variables；
For the unordered classified variable concentrate per a kind of unordered classified variable, count in such unordered classified variable because
Variable-value is classification accounting of the unordered classified variable of target classification value in two classified variables in such unordered classified variable；
Based on the classification accounting of all kinds of unordered classified variables, clustering processing is carried out to the unordered classified variable collection, is obtained
Multiple unordered classified variable subsets；Wherein, each unordered classified variable subset includes at least a kind of unordered classified variable, and each
Unordered classified variable subset is a corresponding orderly classified variable.
With reference to first aspect, the embodiment of the present application provides the first possible embodiment of first aspect, wherein institute
The classification accounting based on all kinds of unordered classified variables is stated, clustering processing is carried out to the unordered classified variable collection, obtains multiple nothings
Sequence classified variable subset, including：
The barycenter for randomly selecting out preset quantity classification as cluster is concentrated from the unordered classified variable；
The unordered classified variable is concentrated in the cluster corresponding to remaining classification distribution to the minimum barycenter of distance；Its
In, the distance between the remaining classification and each barycenter are determined by classification accounting between the two；
The barycenter of each cluster is recalculated, and based on the barycenter after calculating, the unordered classified variable is concentrated again
Each classification carry out cluster distribution, until judge updated barycenter with update before barycenter meet pre-determined distance threshold value
When, stop cluster distribution, obtains multiple unordered classified variable subsets after clustering processing.
With reference to first aspect, the embodiment of the present application provides second of possible embodiment of first aspect, wherein institute
The method of stating further includes：
Determine the classification assignment information of each unordered classified variable subset.
Second of possible embodiment with reference to first aspect, the embodiment of the present application provide the third of first aspect
Possible embodiment, wherein before the classification assignment information for determining each unordered classified variable subset, the method is also wrapped
It includes：
According to the sequence that the classification accounting of unordered classified variable subset is ascending, to each unordered classified variable subset into
The classification accounting of row sequence, the unordered classified variable subset is become by unordered classification of respectively classifying in the unordered classified variable subset
The classification accounting of amount determines；
The classification assignment information of each unordered classified variable subset of the determination, including：
To each unordered classified variable subset carry out sequence assignment after sequence, each unordered classified variable subset is obtained
Classification assignment information.
Second of possible embodiment with reference to first aspect, the embodiment of the present application provide the 4th kind of first aspect
Possible embodiment, wherein the classification assignment information of each unordered classified variable subset of the determination, including：
For each unordered classified variable subset, it is target classification value in two classified variables to calculate the dependent variable value
Unordered classified variable the unordered classified variable subset first classify accounting, and, the dependent variable value be two classification become
The unordered classified variable of the unordered classified variable of non-targeted classification value is accounted in the second classification of the unordered classified variable subset in amount
Than calculating the ratio of the first classification accounting and the second classification accounting, obtaining the first ratio；
For each unordered classified variable subset, it is target classification value in two classified variables to calculate the dependent variable value
Unordered classified variable the multiple unordered classified variable subset third classify accounting, and, the dependent variable value be two
In classified variable the unordered classified variable of non-targeted classification value the multiple unordered classified variable subset the 4th classify accounting,
The ratio for calculating third classification accounting and the 4th classification accounting, obtains the second ratio；
Based on first ratio and second ratio, the classification assignment letter of each unordered classified variable subset is determined
Breath.
Second of possible embodiment with reference to first aspect, the embodiment of the present application provide the 5th kind of first aspect
Possible embodiment, wherein the unordered classified variable is the disorder feature variable in default disaggregated model；
Unordered classified variable collection is obtained, including：
Obtain disorder feature variable；
For the unordered classified variable concentrate per a kind of unordered classified variable, count in such unordered classified variable because
Variable-value is classification accounting of the unordered classified variable of target classification value in two classified variables in such unordered classified variable,
Including：
For, per a kind of disorder feature variable, counted in the disorder feature variables set in such disorder feature variable because
Variable-value is classification accounting of the disorder feature variable of target classification value in two classified variables in such disorder feature variable；
Based on the classification accounting of all kinds of unordered classified variables, clustering processing is carried out to the unordered classified variable collection, is obtained
Multiple unordered classified variable subsets, wherein each unordered classified variable subset includes at least a kind of unordered classified variable, and each
Unordered classified variable subset is a corresponding orderly classified variable, including：
Based on the classification accounting of all kinds of disorder feature variables, clustering processing is carried out to the disorder feature variables set, is obtained
Multiple disorder feature variable subsets, wherein each disorder feature variable subset includes at least a kind of disorder feature variable, and each
Disorder feature variable subset is a corresponding order characteristics variable.
The 5th kind of possible embodiment with reference to first aspect, the embodiment of the present application provide the 6th kind of first aspect
Possible embodiment, wherein after the classification assignment information for determining each unordered classified variable subset, the method is also wrapped
It includes：
Classification assignment information is determined as to the characteristic value of corresponding order characteristics variable；
The argument value that independent variable is corresponded to using the characteristic value as the default disaggregated model inputs the default classification
Model carries out model training.
Second aspect, the embodiment of the present application also provides a kind of unordered classified variable processing unit, described device includes：
Classified variable collection acquisition module, for obtaining unordered classified variable collection, wherein the unordered classified variable collection includes
The unordered classified variable of at least two classes, and corresponding dependent variable is two classified variables；
Classification accounting statistical module, every unordered classified variable of one kind for being concentrated for the unordered classified variable, system
Unordered classified variable that dependent variable value in such unordered classified variable is target classification value in two classified variables is counted in such nothing
Classification accounting in sequence classified variable；
Clustering processing module is used for the classification accounting based on all kinds of unordered classified variables, to the unordered classified variable collection
Clustering processing is carried out, multiple unordered classified variable subsets are obtained；Wherein, each unordered classified variable subset includes at least a kind of nothing
Sequence classified variable, and each unordered classified variable subset is a corresponding orderly classified variable.
In conjunction with second aspect, the embodiment of the present application provides the first possible embodiment of second aspect, wherein institute
Clustering processing module is stated, is specifically used for：
The barycenter for randomly selecting out preset quantity classification as cluster is concentrated from the unordered classified variable；
The unordered classified variable is concentrated in the cluster corresponding to remaining classification distribution to the minimum barycenter of distance；Its
In, the distance between the remaining classification and each barycenter are determined by classification accounting between the two；
The barycenter of each cluster is recalculated, and based on the barycenter after calculating, the unordered classified variable is concentrated again
Each classification carry out cluster distribution, until judge updated barycenter with update before barycenter meet pre-determined distance threshold value
When, stop cluster distribution, obtains multiple unordered classified variable subsets after clustering processing.
In conjunction with second aspect, the embodiment of the present application provides second of possible embodiment of second aspect, wherein institute
Stating device further includes：
Assignment information determining module, the classification assignment information for determining each unordered classified variable subset.
Unordered classified variable processing method and processing device provided by the embodiments of the present application, obtains unordered classified variable collection first,
Wherein, the unordered classified variable collection includes the unordered classified variable of at least two classes, and corresponding dependent variable is two classified variables；Then
For every a kind of unordered classified variable that the unordered classified variable is concentrated, dependent variable value in such unordered classified variable is counted
For classification accounting of the unordered classified variable in such unordered classified variable of target classification value in two classified variables；It is finally based on
The classification accounting of all kinds of unordered classified variables carries out clustering processing to the unordered classified variable collection, obtains multiple unordered classification
Variable subset；Wherein, each unordered classified variable subset includes at least a kind of unordered classified variable, and each unordered classified variable
Subset is a corresponding orderly classified variable, based on the corresponding dependent variable of unordered classified variable in every unordered classified variable of one kind
In classification accounting, clustering processing is carried out to unordered classified variable collection, with multiple unordered classified variable subsets after being grouped,
Grouping can be realized in participation without artificial experience so that the efficiency of packet transaction is higher, and further improves group result
Objectivity and accuracy.
To enable the above objects, features, and advantages of the application to be clearer and more comprehensible, preferred embodiment cited below particularly, and coordinate
Appended attached drawing, is described in detail below.
Description of the drawings
It, below will be to needed in the embodiment attached in order to illustrate more clearly of the technical solution of the embodiment of the present application
Figure is briefly described, it should be understood that the following drawings illustrates only some embodiments of the application, therefore is not construed as pair
The restriction of range for those of ordinary skill in the art without creative efforts, can also be according to this
A little attached drawings obtain other relevant attached drawings.
Fig. 1 shows a kind of flow chart for unordered classified variable processing method that the embodiment of the present application is provided；
Fig. 2 shows the flow charts for the unordered classified variable processing method of another kind that the embodiment of the present application is provided；
Fig. 3 shows the flow chart for the unordered classified variable processing method of another kind that the embodiment of the present application is provided；
Fig. 4 shows the flow chart for the unordered classified variable processing method of another kind that the embodiment of the present application is provided；
Fig. 5 shows the flow chart for the unordered classified variable processing method of another kind that the embodiment of the present application is provided；
Fig. 6 shows the flow chart for the unordered classified variable processing method of another kind that the embodiment of the present application is provided；
Fig. 7 shows a kind of structural schematic diagram for unordered classified variable processing unit that the embodiment of the present application is provided；
Fig. 8 shows a kind of structural schematic diagram for computer equipment that the embodiment of the present application is provided.
Specific implementation mode
To keep the purpose, technical scheme and advantage of the embodiment of the present application clearer, below in conjunction with the embodiment of the present application
Middle attached drawing, technical solutions in the embodiments of the present application are clearly and completely described, it is clear that described embodiment is only
It is some embodiments of the present application, instead of all the embodiments.The application being usually described and illustrated herein in the accompanying drawings is real
Applying the component of example can be arranged and designed with a variety of different configurations.Therefore, below to the application's for providing in the accompanying drawings
The detailed description of embodiment is not intended to limit claimed scope of the present application, but is merely representative of the selected reality of the application
Apply example.Based on embodiments herein, institute that those skilled in the art are obtained without making creative work
There is other embodiment, shall fall in the protection scope of this application.
In view of the method being rule of thumb grouped in the related technology is less efficient, and it cannot ensure the effective of grouping
Property, without being grouped the method directly used then for the relatively broad more classified variables of classifying, cannot ensure follow-up
The effect when applications such as modeling is based on this, and the embodiment of the present application provides a kind of unordered classified variable processing method and processing device,
While raising is grouped the efficiency of processing to unordered classified variable, the objectivity and accuracy of group result are also improved.
As shown in Figure 1, for the flow chart of unordered classified variable processing method provided by the embodiments of the present application, the unordered classification
The executive agent of variable processing method can be computer equipment, and the above method specifically comprises the following steps：
S101, unordered classified variable collection is obtained, wherein unordered classified variable collection includes the unordered classified variable of at least two classes,
And corresponding dependent variable is two classified variables.
Here, unordered classified variable collection can be made of the unordered classified variable of at least two classes.Wherein, unordered classified variable
(unordered categorical variable) refer to the difference without degree and sequence between sub-category or attribute, can be with
It is divided into binomial classification and multinomial classification.If gender (man, female) belongs to binomial classification, and blood group (O, A, B, AB) belongs to multinomial
Classification.Unordered classified variable processing method provided by the embodiments of the present application is applicable not only to binomial classification, applies also for multinomial point
Class.
In order to handle above-mentioned unordered classified variable, the embodiment of the present application also obtains corresponding with unordered classified variable
Two classified variables as dependent variable, it is expected under the supervision of two classified variable to above-mentioned unordered classified variable
Reason promotes the objectivity and accuracy of variable processing.
It is worth noting that unordered classified variable processing method provided by the embodiments of the present application can be applied to various scenes
In, such as medical diagnosis on disease field, character loan field or other have the field that unordered classified variable and dependent variable participate in.For
Convenient for being explained further above-mentioned unordered classified variable processing method, next carried out specifically by taking character loan field as an example
It is bright.Unordered classified variable collection instruction be the academic information of multi-user when, the unordered classified variable of corresponding at least two classes can
To be junior middle school, senior middle school, undergraduate course, postgraduate etc., and corresponding to that can be corresponding with multiple users per the unordered classified variable of class, two classify
Whether variable can then be chosen for overdue.
S102, the every a kind of unordered classified variable concentrated for unordered classified variable, count in such unordered classified variable
Dependent variable value is that classification of the unordered classified variable of target classification value in two classified variables in such unordered classified variable accounts for
Than.
Here, target classification value can be a predetermined threshold value, can also be the classification results determined according to predetermined threshold value,
Such as this application scenarios of above-mentioned character loan, which can be one and preset the overdue amount of money, can also be pair
Overdue classification results are answered, in this way, for above-mentioned per a kind of unordered classified variable (such as senior middle school), if senior middle school's educational background is right
The user answered is 50 people, in 50 people, has the overdue refund of 20 people, the not overdue refund of 30 people, then this kind of unordered classification of senior middle school to become
It is 20/50 that dependent variable value, which is the classification accounting of overdue unordered classified variable, in amount, i.e., 40%.Similarly, for other classes
Unordered classified variable can determine that dependent variable value is overdue unordered classified variable in the unordered classified variable of other classes respectively
Classification accounting, in this way, having obtained corresponding point per a kind of unordered classified variable for what entire unordered classified variable was concentrated
Class accounting.
S103, the classification accounting based on all kinds of unordered classified variables carry out clustering processing to unordered classified variable collection, obtain
Multiple unordered classified variable subsets；Wherein, each unordered classified variable subset includes at least a kind of unordered classified variable, and each
Unordered classified variable subset is a corresponding orderly classified variable.
Here, in order to preferably be grouped processing to above-mentioned all kinds of unordered classified variables, the embodiment of the present application is based on each
The classification accounting of the unordered classified variable of class has carried out clustering processing, with obtain belonging to a kind of unordered classified variable formed it is each
A unordered classified variable subset.In this way, the classified variable of unordered type after cluster, will be grouped into corresponding ordered categorization change
Amount, and each correspond to orderly classified variable and corresponded with each unordered classified variable subset, consequently facilitating after using grouping
Ordered categorization variable carries out subsequent processing, such as classification prediction.
The classification accounting of all kinds of unordered classified variables of the embodiment of the present application based on statistics carries out unordered classified variable collection
The unordered classified variable for belonging to a kind of is condensed together, obtains each unordered classified variable subset by clustering processing.Such as Fig. 2 institutes
Show, above-mentioned cluster process is realized especially by following steps：
S201, the barycenter for randomly selecting out preset quantity classification as cluster is concentrated from unordered classified variable；
S202, concentrate remaining classification distribution in the cluster corresponding to minimum barycenter unordered classified variable；
Wherein, the distance between remaining classification and each barycenter are determined by classification accounting between the two；
S203, the barycenter for recalculating each cluster, and based on the barycenter after calculating, unordered classified variable is concentrated again
Each classification carry out cluster distribution, until judge updated barycenter with update before barycenter meet pre-determined distance threshold value
When, stop cluster distribution, obtains multiple unordered classified variable subsets after clustering processing.
Here, the embodiment of the present application randomly selects preset quantity (such as K) classification for unordered classified variable collection, as
The initial barycenter of cluster, and each classification is corresponding with the classification accounting of the unordered classified variable of every class, in this way, except by K
After other corresponding classification accountings of classifying are compared with corresponding classification accountings of each classification in K classification, determination and
Each most similar cluster of classification in other classification, and other each classification in classifying are distributed into determining cluster, it is first
Wheel cluster is completed.After the completion of the first run clusters, the barycenter currently clustered can also be redefined, is determining the barycenter currently clustered
Afterwards, then again each classification concentrated to unordered classified variable carries out cluster distribution, until judge updated barycenter with
When barycenter before update meets pre-determined distance threshold value, stop cluster distribution, that is, cluster is with the most phase of the classification in each cluster
Closely, classification between clustering is farthest away from for stop condition, to obtain multiple unordered classified variable subsets after clustering processing.
It is worth noting that K-Means clustering methods, which may be used, in the embodiment of the present application realizes above-mentioned processing procedure, may be used also
To realize that above-mentioned processing procedure, comparison do not do specific limitation using other clustering methods.
In view of unordered classified variable processing method provided by the embodiments of the present application can be adapted in character loan field
Classification prediction in, therefore, the embodiment of the present application also can determine whether the classification assignment information of each unordered classified variable subset, with root
The training for carrying out model as the input feature vector of disaggregated model according to the unordered classified variable after classification assignment, to improve above-mentioned nothing
The applicability of sequence classified variable processing method.
Unordered classified variable processing method provided by the embodiments of the present application can carry out classification assignment in the following way：The
A kind of mode is first to be ranked up unordered classified variable subset, then to the unordered classified variable subset carry out sequence tax after sequence
Value, another way are a kind of assignment modes using direct categorization values, can also be the combination side of above two mode
Formula is first ranked up, then carry out categorization values processing to carry out classification assignment.Next it is divided to two aspects to above two
Mode is specifically described.
First aspect：As shown in figure 3, above-mentioned classification assignment information is determined as follows：
S301, according to the ascending sequence of the classification accounting of unordered classified variable subset, to each unordered classified variable
Subset is ranked up, and the classification accounting of unordered classified variable subset is by unordered classified variable of respectively classifying in unordered classified variable subset
Classification accounting determine；
S302, to each unordered classified variable subset carry out sequence assignment after sequence, obtain each unordered classified variable
The classification assignment information of subset.
Here, after clustering processing, each unordered classified variable subset includes at least a kind of unordered classified variable, the nothing
The classification accounting of sequence classified variable subset then can be by the classification for unordered classified variable of respectively classifying in the unordered classified variable subset
Accounting determines, can be determined by the maximum value of the classification accounting for unordered classified variable of respectively classifying in unordered classified variable subset, also
It can be determined by the minimum value of the classification accounting for unordered classified variable of respectively classifying in unordered classified variable subset, it can also be by other
Unordered classified variable subset can be determined in such a way that classification accounting is ranked up, the embodiment of the present application, which does not do this, to be had
The limitation of body.
In addition, to each unordered classified variable subset carry out sequence assignment after sequence, such as assigned according to 1,2,3 ...
Value, can obtain the classification assignment information of each unordered classified variable subset.
Second aspect：As shown in figure 4, above-mentioned classification assignment information is determined as follows：
S401, it is directed to each unordered classified variable subset, it is target classification value in two classified variables to calculate dependent variable value
Unordered classified variable the unordered classified variable subset first classify accounting, and, dependent variable value be two classified variables
In non-targeted classification value unordered classified variable unordered classified variable the unordered classified variable subset second classify accounting,
The ratio for calculating the first classification accounting and the second classification accounting, obtains the first ratio；
S402, it is directed to each unordered classified variable subset, it is target classification value in two classified variables to calculate dependent variable value
Unordered classified variable multiple unordered classified variable subsets third classify accounting, and, dependent variable value be two classification become
The unordered classified variable of non-targeted classification value calculates third point in the 4th classification accounting of multiple unordered classified variable subsets in amount
The ratio of class accounting and the 4th classification accounting, obtains the second ratio；
S403, it is based on the first ratio and the second ratio, determines the classification assignment information of each unordered classified variable subset.
Here, after clustering processing, each unordered classified variable subset includes at least a kind of unordered classified variable, for
Each unordered classified variable subset calculates the first classification accounting, the second classification accounting to determine the first ratio, also calculates third point
Class accounting, the 4th classification accounting are finally based on the first ratio and the second ratio to determine the second ratio, determine each unordered classification
The classification assignment information of variable subset.
Such as this application scenarios of above-mentioned character loan, for each unordered classified variable subset, which accounts for
It is that overdue all kinds of unordered classified variables are accounted in the classification of the unordered classified variable subset than that can be according to dependent variable value
Ratio determines that it is not overdue all kinds of unordered points that the second classification accounting, which can be then according to dependent variable value, with Data-Statistics result
Class variable the unordered classified variable subset classification accounting and Data-Statistics result determine, according to the first classification accounting with second
The ratio for accounting of classifying, you can obtain the first ratio Py1.In addition, above-mentioned third classification accounting can be according to dependent variable value
Classification accounting of the overdue all kinds of unordered classified variables in unordered classified variable collection (corresponding multiple unordered classified variable subsets)
And Data-Statistics result determine, it is above-mentioned 4th classification accounting can be according to dependent variable value be not overdue all kinds of unordered classification
Variable unordered classified variable collection (corresponding multiple unordered classified variable subsets) classification accounting and the determination of Data-Statistics result, root
According to the ratio of third classification accounting and the 4th classification accounting, you can obtain the second ratio Py0.The embodiment of the present application can be based on down
Formula determines the classification assignment information of each unordered classified variable subset：
Based on above formula it is found that in the embodiment of the present application classification assignment information reflection be to exceed in the case where independent variable is each grouped
Phase user is to difference of the overdue user between normal users accounting in normal users accounting and totality, that is, above-mentioned classification is assigned
Value information has contained influence of the independent variable value for target dependent variable (overdue probability), consequently facilitating being believed using category assignment
Breath establishes generalized linear model to carry out the operations such as classification prediction.
With reference to first aspect and second aspect, the embodiment of the present application also provides a kind of classification assignment informations to determine method,
Specifically comprise the following steps：
Step 1: according to the ascending sequence of the classification accounting of unordered classified variable subset, each unordered classification is become
Quantum collection is ranked up, and the classification accounting of unordered classified variable subset is become by unordered classification of respectively classifying in unordered classified variable subset
The classification accounting of amount determines；
Step 2: for each unordered classified variable subset, it is target classification in two classified variables to calculate dependent variable value
Value unordered classified variable the unordered classified variable subset first classify accounting, and, dependent variable value be two classification become
The unordered classified variable of the unordered classified variable of non-targeted classification value is accounted in the second classification of the unordered classified variable subset in amount
Than calculating the ratio of the first classification accounting and the second classification accounting, obtaining the first ratio；
Step 3: for each unordered classified variable subset, it is target classification in two classified variables to calculate dependent variable value
Value unordered classified variable multiple unordered classified variable subsets third classify accounting, and, dependent variable value be two classification
The unordered classified variable of non-targeted classification value calculates third in the 4th classification accounting of multiple unordered classified variable subsets in variable
The ratio of accounting of classifying and the 4th classification accounting, obtains the second ratio；
Step 4: it is based on the first ratio and the second ratio, the classification of unordered classified variable subset each of after determining sequence
Assignment information.
The specific implementation process of above-mentioned steps is not done superfluous herein referring to the detailed description to first aspect and second aspect
It states.
In view of unordered classified variable processing method provided by the embodiments of the present application can be applied to the feature of model training
In variable processing procedure, therefore, in the disorder feature variable during unordered classified variable is default disaggregated model, as shown in figure 5,
Above-mentioned unordered classified variable processing method is realized especially by following steps：
S501, disorder feature variable is obtained；
S502, for, per a kind of disorder feature variable, being counted in such disorder feature variable in disorder feature variables set
Dependent variable value is that classification of the disorder feature variable of target classification value in two classified variables in such disorder feature variable accounts for
Than；
S503, the classification accounting based on all kinds of disorder feature variables carry out clustering processing to disorder feature variables set, obtain
Multiple disorder feature variable subsets, wherein each disorder feature variable subset includes at least a kind of disorder feature variable, and each
Disorder feature variable subset is a corresponding order characteristics variable.
Here, the embodiment of the present application obtains disorder feature variable first, then counts each in disorder feature variables set
Dependent variable value is the disorder feature variable of target classification value in two classified variables in such unordered spy in class disorder feature variable
The classification accounting in variable is levied, the classification accounting of all kinds of disorder feature variables is finally based on, disorder feature variables set is gathered
Class processing, obtains multiple disorder feature variable subsets, wherein each disorder feature variable subset includes at least a kind of disorder feature
Variable, and each disorder feature variable subset is a corresponding order characteristics variable.
Likewise, the embodiment of the present application can also be to multiple unordered classified variables that above-mentioned disorder feature Variable cluster obtains
Subset carries out classification assignment respectively, after classification assignment, can carry out model training.As shown in fig. 6, above-mentioned model training mistake
Journey specifically comprises the following steps：
S601, the characteristic value that classification assignment information is determined as to corresponding order characteristics variable；
S602, the argument value input that independent variable is corresponded to using characteristic value as default disaggregated model are preset disaggregated model and are carried out
Model training.
Here, it is contemplated that (such as logistic is returned the assignment mode of direct categorization values with some generalized linear models
Model) in target variable logistic conversion regimes correspondence that is consistent, therefore can directly determining classification assignment information
The characteristic value of order characteristics variable corresponds to the argument value of independent variable as default disaggregated model, to participate in model training, is applicable in
Property is stronger.
For the ease of further understanding unordered classified variable processing method provided by the embodiments of the present application, existing special credit of lifting is borrowed
This example is borrowed to illustrate.
Table 1
As shown in table 1, it is concentrated including the unordered classified variable that user is junior middle school, senior middle school, junior college, undergraduate course, postgraduate etc.,
The unordered classified variable of corresponding 8 classes, whether corresponding two classified variable can be chosen overdue.For this classification of junior middle school, 50
People is overdue, and 200 people are not overdue, then dependent variable value is that overdue unordered classification becomes in this kind of unordered classified variable of junior middle school
The classification accounting of amount be 50/ (50+200), i.e., 0.2, similarly, for this classification of senior middle school, corresponding classification accounting is
20/ (20+200), i.e., 0.09, for this classification of junior college, corresponding classification accounting be 5/ (5+200), i.e., 0.024,
For undergraduate course this classification, corresponding classification accounting is 15/ (15+200), i.e., 0.07, for this classification of postgraduate
For, corresponding classification accounting be 10/ (10+200), i.e., 0.05.In this way, it is based on above-mentioned each classification accounting, it can will be upper
Unordered classified variable clustering processing is stated, 3 unordered classified variable subsets, the respectively first unordered classified variable subset are obtained
{ junior college, postgraduate }, the second unordered classified variable subset { undergraduate course, senior middle school }, the unordered classified variable subset { junior middle school } of third.For
Each unordered classified variable subset can carry out classification assignment using formula (1).Wherein, the first unordered classified variable subset
Corresponding first ratio Py1For (5+10)/(200+200)=15/400, the second ratio Py0Be 100/1000, then WOE1=ln
((15/400)/(100/1000))=- 0.981, the corresponding first ratio P of the second unordered classified variable subsety1For (20+15)/
(200+200)=35/400, the second ratio Py0Be 100/1000, then WOE2=ln ((35/400)/(100/1000))=-
0.223, the corresponding first ratio P of the unordered classified variable subset of thirdy1It is 50/200, the second ratio Py0It is 100/1000, then
WOE3=ln ((50/200)/(100/1000))=0.916, then, junior college, postgraduate in the first unordered classified variable subset
This unordered classified variable of two classes has been clustered into the first ordered categorization variable, and class label is -0.981, the second unordered classified variable
Undergraduate course, senior middle school this unordered classified variable of two classes in subset has been clustered into the second ordered categorization variable, and class label is -0.223,
This kind of unordered classified variable in junior middle school in the unordered classified variable subset of third has been clustered into third ordered categorization variable, and class label
It is 0.916.To convert unordered classified variable for ordered categorization variable, transformed ordered categorization variable can directly join
With the training of disaggregated model, applicability is stronger.
Based on same inventive concept, nothing corresponding with unordered classified variable processing method is additionally provided in the embodiment of the present application
Sequence classified variable processing unit, the principle and the above-mentioned nothing of the embodiment of the present application solved the problems, such as due to the device in the embodiment of the present application
Sequence classified variable processing method is similar, therefore the implementation of device may refer to the implementation of method, and overlaps will not be repeated.Such as figure
Shown in 7, by the structural schematic diagram for the unordered classified variable processing unit that the embodiment of the present application provides, at the unordered classified variable
Managing device includes：
Classified variable collection acquisition module 701, for obtaining unordered classified variable collection, wherein unordered classified variable collection includes
The unordered classified variable of at least two classes, and corresponding dependent variable is two classified variables；
Accounting of classifying statistical module 702, every a kind of unordered classified variable for being concentrated for unordered classified variable, statistics
Dependent variable value is that the unordered classified variable of target classification value in two classified variables is unordered at such in such unordered classified variable
Classification accounting in classified variable；
Clustering processing module 703, be used for the classification accounting based on all kinds of unordered classified variables, to unordered classified variable collection into
Row clustering processing obtains multiple unordered classified variable subsets；Wherein, each unordered classified variable subset includes at least a kind of unordered
Classified variable, and each unordered classified variable subset is a corresponding orderly classified variable.
In one embodiment, above-mentioned clustering processing module 703, is specifically used for：
The barycenter for randomly selecting out preset quantity classification as cluster is concentrated from unordered classified variable；
Unordered classified variable is concentrated in the cluster corresponding to remaining classification distribution to the minimum barycenter of distance；Wherein,
The distance between remaining classification and each barycenter are determined by classification accounting between the two；
The barycenter of each cluster is recalculated, and based on the barycenter after calculating, unordered classified variable is concentrated again every
One classification carries out cluster distribution, until when the barycenter before judging updated barycenter and update meets pre-determined distance threshold value,
Stop cluster distribution, obtains multiple unordered classified variable subsets after clustering processing.
In another embodiment, above-mentioned unordered classified variable processing unit further includes：
Assignment information determining module 704, the classification assignment information for determining each unordered classified variable subset.
In another embodiment, above-mentioned unordered classified variable processing unit further includes：
Classified variable subset sort module 705, for ascending according to the classification accounting of unordered classified variable subset
Sequentially, each unordered classified variable subset is ranked up, the classification accounting of unordered classified variable subset is by unordered classified variable
Respectively classify in subset unordered classified variable classification accounting determine；
Assignment information determining module 704 is specifically used for each unordered classified variable subset carry out sequence tax after sequence
Value, obtains the classification assignment information of each unordered classified variable subset.
In another embodiment, above-mentioned assignment information determining module 704 is specifically used for：
For each unordered classified variable subset, it is the unordered of target classification value in two classified variables to calculate dependent variable value
Classified variable the unordered classified variable subset first classify accounting, and, dependent variable value be two classified variables in non-mesh
It marks the unordered classified variable of the unordered classified variable of classification value the second of the unordered classified variable subset to classify accounting, calculates the
The ratio of one classification accounting and the second classification accounting, obtains the first ratio；
For each unordered classified variable subset, it is the unordered of target classification value in two classified variables to calculate dependent variable value
Classified variable multiple unordered classified variable subsets third classify accounting, and, dependent variable value be two classified variables in it is non-
The unordered classified variable of target classification value calculates third classification accounting in the 4th classification accounting of multiple unordered classified variable subsets
With the ratio of the 4th classification accounting, the second ratio is obtained；
Based on the first ratio and the second ratio, the classification assignment information of each unordered classified variable subset is determined.
In another embodiment, unordered classified variable is the disorder feature variable in default disaggregated model；
Classified variable collection acquisition module 701 is specifically used for obtaining disorder feature variable；
Accounting of classifying statistical module 702 is specifically used for for every a kind of disorder feature variable in disorder feature variables set,
Disorder feature variable that dependent variable value in such disorder feature variable is target classification value in two classified variables is counted at such
Classification accounting in disorder feature variable；
Clustering processing module 703 is specifically used for the classification accounting based on all kinds of disorder feature variables, to disorder feature variable
Collection carries out clustering processing, obtains multiple disorder feature variable subsets, wherein each disorder feature variable subset includes at least a kind of
Disorder feature variable, and each disorder feature variable subset is a corresponding order characteristics variable.
In another embodiment, above-mentioned unordered classified variable processing unit further includes：
Model training module 706, the characteristic value for classification assignment information to be determined as to corresponding order characteristics variable；It will be special
Disaggregated model progress model training is preset in the argument value input that value indicative corresponds to independent variable as default disaggregated model.
As shown in figure 8, a kind of structural schematic diagram of the computer equipment provided by the embodiment of the present application, the computer are set
It is standby to include：Processor 801, memory 802 and bus 803, the storage of memory 802 executes instruction, when device is run, processor
It is communicated by bus 803 between 801 and memory 802, what is stored in the execution memory 802 of processor 801 executes instruction as follows：
Obtain unordered classified variable collection, wherein unordered classified variable collection includes the unordered classified variable of at least two classes, and corresponding
Dependent variable is two classified variables；
For every a kind of unordered classified variable that unordered classified variable is concentrated, dependent variable in such unordered classified variable is counted
Value is classification accounting of the unordered classified variable of target classification value in two classified variables in such unordered classified variable；
Based on the classification accounting of all kinds of unordered classified variables, clustering processing is carried out to unordered classified variable collection, is obtained multiple
Unordered classified variable subset；Wherein, each unordered classified variable subset includes at least a kind of unordered classified variable, and each unordered
Classified variable subset is a corresponding orderly classified variable.
In one embodiment, in the processing that above-mentioned processor 801 executes, the classification based on all kinds of unordered classified variables
Accounting carries out clustering processing to unordered classified variable collection, obtains multiple unordered classified variable subsets, including：
The barycenter for randomly selecting out preset quantity classification as cluster is concentrated from unordered classified variable；
Unordered classified variable is concentrated in the cluster corresponding to remaining classification distribution to the minimum barycenter of distance；Wherein,
The distance between remaining classification and each barycenter are determined by classification accounting between the two；
The barycenter of each cluster is recalculated, and based on the barycenter after calculating, unordered classified variable is concentrated again every
One classification carries out cluster distribution, until when the barycenter before judging updated barycenter and update meets pre-determined distance threshold value,
Stop cluster distribution, obtains multiple unordered classified variable subsets after clustering processing.
In another embodiment, above-mentioned processor 801 is additionally operable to determine the classification of each unordered classified variable subset
Assignment information.
In another embodiment, in the processing that above-mentioned processor 801 executes, each unordered classified variable is being determined
Before the classification assignment information of collection, it is additionally operable to：
According to the sequence that the classification accounting of unordered classified variable subset is ascending, to each unordered classified variable subset into
Row sequence, the classification accounting of unordered classified variable subset by unordered classified variable of respectively classifying in unordered classified variable subset classification
Accounting determines；
Determine the classification assignment information of each unordered classified variable subset, including：
To each unordered classified variable subset carry out sequence assignment after sequence, each unordered classified variable subset is obtained
Classification assignment information.
In another embodiment, in the processing that above-mentioned processor 801 executes, each unordered classified variable subset is determined
Classification assignment information, including：
For each unordered classified variable subset, it is the unordered of target classification value in two classified variables to calculate dependent variable value
Classified variable the unordered classified variable subset first classify accounting, and, dependent variable value be two classified variables in non-mesh
It marks the unordered classified variable of the unordered classified variable of classification value the second of the unordered classified variable subset to classify accounting, calculates the
The ratio of one classification accounting and the second classification accounting, obtains the first ratio；
For each unordered classified variable subset, it is the unordered of target classification value in two classified variables to calculate dependent variable value
Classified variable multiple unordered classified variable subsets third classify accounting, and, dependent variable value be two classified variables in it is non-
The unordered classified variable of target classification value calculates third classification accounting in the 4th classification accounting of multiple unordered classified variable subsets
With the ratio of the 4th classification accounting, the second ratio is obtained；
Based on the first ratio and the second ratio, the classification assignment information of each unordered classified variable subset is determined.
In another embodiment, unordered classified variable is the disorder feature variable in default disaggregated model；Above-mentioned place
It manages in the processing that device 801 executes：
Unordered classified variable collection is obtained, including：
Obtain disorder feature variable；
For every a kind of unordered classified variable that unordered classified variable is concentrated, dependent variable in such unordered classified variable is counted
Value is classification accounting of the unordered classified variable of target classification value in two classified variables in such unordered classified variable, packet
It includes：
For every a kind of disorder feature variable in disorder feature variables set, dependent variable in such disorder feature variable is counted
Value is classification accounting of the disorder feature variable of target classification value in two classified variables in such disorder feature variable；
Based on the classification accounting of all kinds of unordered classified variables, clustering processing is carried out to unordered classified variable collection, is obtained multiple
Unordered classified variable subset, wherein each unordered classified variable subset includes at least a kind of unordered classified variable, and each unordered
Classified variable subset is a corresponding orderly classified variable, including：
Based on the classification accounting of all kinds of disorder feature variables, clustering processing is carried out to disorder feature variables set, is obtained multiple
Disorder feature variable subset, wherein each disorder feature variable subset includes at least a kind of disorder feature variable, and each unordered
Characteristic variable subset is a corresponding order characteristics variable.
In another embodiment, in the processing that above-mentioned processor 801 executes, each unordered classified variable is being determined
After the classification assignment information of collection, it is additionally operable to：
Classification assignment information is determined as to the characteristic value of corresponding order characteristics variable；
Disaggregated model progress model is preset in the argument value input that independent variable is corresponded to using characteristic value as default disaggregated model
Training.
The embodiment of the present application also provides a kind of computer readable storage medium, stored on the computer readable storage medium
There is computer program, which executes above-mentioned unordered classified variable processing method when being run by processor the step of.
Specifically, which can be general storage medium, such as mobile disk, hard disk, on the storage medium
Computer program when being run, above-mentioned unordered classified variable processing method is able to carry out, to which solve at present can not be to unordered
The problem of classified variable is effectively grouped, and then reach and can improve the efficiency for being grouped processing to unordered classified variable
While, also improve the objectivity of group result and the effect of accuracy.
The computer program product for the unordered classified variable processing method that the embodiment of the present application is provided, including store journey
The computer readable storage medium of sequence code, the instruction that program code includes can be used for executing the side in previous methods embodiment
Method, specific implementation can be found in embodiment of the method, and details are not described herein.
It is apparent to those skilled in the art that for convenience and simplicity of description, the system of foregoing description
It with the specific work process of device, can refer to corresponding processes in the foregoing method embodiment, details are not described herein.
If function is realized in the form of SFU software functional unit and when sold or used as an independent product, can store
In a computer read/write memory medium.Based on this understanding, the technical solution of the application is substantially in other words to existing
There is the part for the part or the technical solution that technology contributes that can be expressed in the form of software products, the computer
Software product is stored in a storage medium, including some instructions are used so that a computer equipment (can be personal meter
Calculation machine, server or network equipment etc.) execute each embodiment method of the application all or part of step.And it is above-mentioned
Storage medium includes：USB flash disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory
The various media that can store program code such as (RAM, Random Access Memory), magnetic disc or CD.
More than, the only specific implementation mode of the application, but the protection domain of the application is not limited thereto, and it is any to be familiar with
Those skilled in the art can easily think of the change or the replacement in the technical scope that the application discloses, and should all cover
Within the protection domain of the application.Therefore, the protection domain of the application should be subject to the protection scope in claims.
Claims (10)
1. a kind of unordered classified variable processing method, which is characterized in that the method includes：
Obtain unordered classified variable collection, wherein the unordered classified variable collection includes the unordered classified variable of at least two classes, and corresponding
Dependent variable is two classified variables；
For every a kind of unordered classified variable that the unordered classified variable is concentrated, dependent variable in such unordered classified variable is counted
Value is classification accounting of the unordered classified variable of target classification value in two classified variables in such unordered classified variable；
Based on the classification accounting of all kinds of unordered classified variables, clustering processing is carried out to the unordered classified variable collection, is obtained multiple
Unordered classified variable subset；Wherein, each unordered classified variable subset includes at least a kind of unordered classified variable, and each unordered
Classified variable subset is a corresponding orderly classified variable.
2. according to the method described in claim 1, it is characterized in that, the classification accounting based on all kinds of unordered classified variables,
Clustering processing is carried out to the unordered classified variable collection, obtains multiple unordered classified variable subsets, including：
The barycenter for randomly selecting out preset quantity classification as cluster is concentrated from the unordered classified variable；
The unordered classified variable is concentrated in the cluster corresponding to remaining classification distribution to the minimum barycenter of distance；Wherein,
The distance between the remaining classification and each barycenter are determined by classification accounting between the two；
The barycenter of each cluster is recalculated, and based on the barycenter after calculating, the unordered classified variable is concentrated again every
One classification carries out cluster distribution, until when the barycenter before judging updated barycenter and update meets pre-determined distance threshold value,
Stop cluster distribution, obtains multiple unordered classified variable subsets after clustering processing.
3. according to the method described in claim 1, it is characterized in that, the method further includes：
Determine the classification assignment information of each unordered classified variable subset.
4. according to the method described in claim 3, it is characterized in that, in the classification assignment for determining each unordered classified variable subset
Before information, the method further includes：
According to the sequence that the classification accounting of unordered classified variable subset is ascending, each unordered classified variable subset is arranged
Sequence, the classification accounting of the unordered classified variable subset is by unordered classified variable of respectively classifying in the unordered classified variable subset
Accounting of classifying determines；
The classification assignment information of each unordered classified variable subset of the determination, including：
To each unordered classified variable subset carry out sequence assignment after sequence, the classification of each unordered classified variable subset is obtained
Assignment information.
5. according to the method described in claim 3, it is characterized in that, the determination each unordered classified variable subset classification assign
Value information, including：
For each unordered classified variable subset, it is the unordered of target classification value in two classified variables to calculate the dependent variable value
Classified variable the unordered classified variable subset first classify accounting, and, the dependent variable value be two classified variables in
The unordered classified variable of the unordered classified variable of non-targeted classification value is in the second classification accounting of the unordered classified variable subset, meter
The ratio for calculating the first classification accounting and the second classification accounting, obtains the first ratio；
For each unordered classified variable subset, it is the unordered of target classification value in two classified variables to calculate the dependent variable value
Classified variable the multiple unordered classified variable subset third classify accounting, and, the dependent variable value be two classification
The unordered classified variable of non-targeted classification value is calculated in the 4th classification accounting of the multiple unordered classified variable subset in variable
The ratio of third classification accounting and the 4th classification accounting, obtains the second ratio；
Based on first ratio and second ratio, the classification assignment information of each unordered classified variable subset is determined.
6. according to the method described in claim 3, it is characterized in that, the unordered classified variable is the nothing in default disaggregated model
Sequence characteristics variable；
Unordered classified variable collection is obtained, including：
Obtain disorder feature variable；
For every a kind of unordered classified variable that the unordered classified variable is concentrated, dependent variable in such unordered classified variable is counted
Value is classification accounting of the unordered classified variable of target classification value in two classified variables in such unordered classified variable, packet
It includes：
For every a kind of disorder feature variable in the disorder feature variables set, dependent variable in such disorder feature variable is counted
Value is classification accounting of the disorder feature variable of target classification value in two classified variables in such disorder feature variable；
Based on the classification accounting of all kinds of unordered classified variables, clustering processing is carried out to the unordered classified variable collection, is obtained multiple
Unordered classified variable subset, wherein each unordered classified variable subset includes at least a kind of unordered classified variable, and each unordered
Classified variable subset is a corresponding orderly classified variable, including：
Based on the classification accounting of all kinds of disorder feature variables, clustering processing is carried out to the disorder feature variables set, is obtained multiple
Disorder feature variable subset, wherein each disorder feature variable subset includes at least a kind of disorder feature variable, and each unordered
Characteristic variable subset is a corresponding order characteristics variable.
7. according to the method described in claim 6, it is characterized in that, in the classification assignment for determining each unordered classified variable subset
After information, the method further includes：
Classification assignment information is determined as to the characteristic value of corresponding order characteristics variable；
The argument value that independent variable is corresponded to using the characteristic value as the default disaggregated model inputs the default disaggregated model
Carry out model training.
8. a kind of unordered classified variable processing unit, which is characterized in that described device includes：
Classified variable collection acquisition module, for obtaining unordered classified variable collection, wherein the unordered classified variable collection includes at least
The unordered classified variable of two classes, and corresponding dependent variable is two classified variables；
Classification accounting statistical module, every unordered classified variable of one kind for being concentrated for the unordered classified variable, statistics should
Dependent variable value is the unordered classified variable of target classification value in two classified variables at such unordered point in the unordered classified variable of class
Classification accounting in class variable；
Clustering processing module is used for the classification accounting based on all kinds of unordered classified variables, is carried out to the unordered classified variable collection
Clustering processing obtains multiple unordered classified variable subsets；Wherein, each unordered classified variable subset includes at least a kind of unordered point
Class variable, and each unordered classified variable subset is a corresponding orderly classified variable.
9. device according to claim 8, which is characterized in that the clustering processing module is specifically used for：
The barycenter for randomly selecting out preset quantity classification as cluster is concentrated from the unordered classified variable；
The unordered classified variable is concentrated in the cluster corresponding to remaining classification distribution to the minimum barycenter of distance；Wherein,
The distance between the remaining classification and each barycenter are determined by classification accounting between the two；
The barycenter of each cluster is recalculated, and based on the barycenter after calculating, the unordered classified variable is concentrated again every
One classification carries out cluster distribution, until when the barycenter before judging updated barycenter and update meets pre-determined distance threshold value,
Stop cluster distribution, obtains multiple unordered classified variable subsets after clustering processing.
10. device according to claim 8, which is characterized in that described device further includes：
Assignment information determining module, the classification assignment information for determining each unordered classified variable subset.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810403750.1A CN108595669A (en) | 2018-04-28 | 2018-04-28 | A kind of unordered classified variable processing method and processing device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810403750.1A CN108595669A (en) | 2018-04-28 | 2018-04-28 | A kind of unordered classified variable processing method and processing device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108595669A true CN108595669A (en) | 2018-09-28 |
Family
ID=63619340
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810403750.1A Pending CN108595669A (en) | 2018-04-28 | 2018-04-28 | A kind of unordered classified variable processing method and processing device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108595669A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110059013A (en) * | 2019-04-24 | 2019-07-26 | 北京百度网讯科技有限公司 | The determination method and device operated normally after software upgrading |
| CN110502469A (en) * | 2019-08-29 | 2019-11-26 | 上海燧原智能科技有限公司 | A kind of data distributing method, device, equipment and storage medium |
-
2018
- 2018-04-28 CN CN201810403750.1A patent/CN108595669A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110059013A (en) * | 2019-04-24 | 2019-07-26 | 北京百度网讯科技有限公司 | The determination method and device operated normally after software upgrading |
| CN110059013B (en) * | 2019-04-24 | 2023-06-23 | 北京百度网讯科技有限公司 | Method and device for determining normal operation after software upgrading |
| CN110502469A (en) * | 2019-08-29 | 2019-11-26 | 上海燧原智能科技有限公司 | A kind of data distributing method, device, equipment and storage medium |
| CN110502469B (en) * | 2019-08-29 | 2020-06-12 | 上海燧原智能科技有限公司 | Data distribution method, device, equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104915879B (en) | The method and device that social relationships based on finance data are excavated | |
| CN107547214B (en) | Group's reading method, electronic equipment and computer storage medium based on e-book | |
| CN108171280A (en) | A kind of grader construction method and the method for prediction classification | |
| CN108182634A (en) | A kind of training method for borrowing or lending money prediction model, debt-credit Forecasting Methodology and device | |
| CN106156092B (en) | Data processing method and device | |
| US10489861B1 (en) | Methods and systems for improving the underwriting process | |
| CN107704995A (en) | Student's evaluation system | |
| US20240202205A1 (en) | Library information management system | |
| CN109389321A (en) | A kind of price evaluation method and device | |
| CN106875185A (en) | A kind of air control model training method and device | |
| CN108595669A (en) | A kind of unordered classified variable processing method and processing device | |
| CN108985600A (en) | Analysis method, computer readable storage medium and the analytical equipment of enterprise competitiveness | |
| CN106897282A (en) | The sorting technique and equipment of a kind of customer group | |
| CN107239853B (en) | Intelligent housekeeper system based on cloud computing and working method thereof | |
| CN118134652A (en) | Asset configuration scheme generation method and device, electronic equipment and medium | |
| CN109785197A (en) | A kind of Learning behavior analyzing method and device | |
| CN105447117A (en) | User clustering method and apparatus | |
| AU2022204475B2 (en) | Generating estimates by combining unsupervised and supervised machine learning | |
| CN110245985A (en) | A kind of information processing method and device | |
| Bhat et al. | Predicting private company exits using qualitative data | |
| CN110428278A (en) | Determine the method and device of resource share | |
| CN112418654B (en) | Resource allocation method and device, electronic equipment and storage medium | |
| CN115167965A (en) | Transaction progress bar processing method and device | |
| CN108629674A (en) | Auction the distribution method and terminal device of income | |
| CN107093149A (en) | Online friend relation strength assessment method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 101-8, 1st floor, building 31, area 1, 188 South Fourth Ring Road West, Fengtai District, Beijing Applicant after: Guoxin Youyi Data Co., Ltd Address before: 100070, No. 188, building 31, headquarters square, South Fourth Ring Road West, Fengtai District, Beijing Applicant before: SIC YOUE DATA Co.,Ltd. |
|
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180928 |