CN108961071A - The method and terminal device of automatic Prediction composite service income - Google Patents
The method and terminal device of automatic Prediction composite service income Download PDFInfo
- Publication number
- CN108961071A CN108961071A CN201810554475.3A CN201810554475A CN108961071A CN 108961071 A CN108961071 A CN 108961071A CN 201810554475 A CN201810554475 A CN 201810554475A CN 108961071 A CN108961071 A CN 108961071A
- Authority
- CN
- China
- Prior art keywords
- sample
- type
- business
- service
- supplementary service
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/08—Insurance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
Abstract
The present invention is suitable for technical field of data processing, provide method, terminal device and the computer readable storage medium of automatic Prediction composite service income, it include: to obtain multiple groups learning sample, and the multiple groups learning sample is handled by preset Processing Algorithm, service selection model is obtained, learning sample described in every group is made of main business sample, supplementary service sample and user's sample；The service selection model is inputted using main business service type and the characteristic value of user object as prediction object information, and is the objective cross business of the user object by the supplementary service type of service selection model output and the main business type combination；From and more set predicting strategies of the main business type association in determine corresponding with objective cross business target prediction strategy, and the financial value of the objective cross business is obtained based on the target prediction strategy.The present invention, which improves, to be predicted objective cross business and carries out forecasting reliability to financial value.
Description
Technical field
The invention belongs to method, the terminals of technical field of data processing more particularly to automatic Prediction composite service income to set
Standby and computer readable storage medium.
Background technique
Business refers to the relevant issues of sale, and income is the interests that business activity can obtain.Before commencing business,
The income of the business would generally be predicted, prediction technique is the correlated characteristic of acquisition business, and fixed according to the business
Calculating logic is calculated.It, can be according to the relevant information of insurer when insurer insures to certain insurance kind to insure citing
Calculate obtainable income of insuring.
Since business activity has the trend that gradually complicates, a kind of composite service may be comprising main business and multinomial additional
Business needs the combination to main business and a certain supplementary service to calculate when calculating income, for example certain insurance includes main danger
With attached danger.In the case where user selectes main business and unselected supplementary service, if calculated using conventional method,
Since supplementary service lacks, financial value can not be calculated, the combination of each supplementary service and main business can only be carried out independent
It calculates, calculated data all do not need very much, and the reliability of earnings forecast is low.
Summary of the invention
In view of this, the embodiment of the invention provides the method for automatic Prediction composite service income, terminal device and calculating
Machine readable storage medium storing program for executing, to solve to carry out the reliable of earnings forecast to composite service in known main business in the prior art
The low problem of property.
The first aspect of the embodiment of the present invention provides a kind of method of automatic Prediction composite service income, comprising:
Multiple groups learning sample is obtained, and the multiple groups learning sample is handled by preset Processing Algorithm, is obtained
Service selection model, learning sample described in every group are made of main business sample, supplementary service sample and user's sample；
The service selection model is inputted using main business service type and the characteristic value of user object as prediction object information, and
It is the target of the user object by the supplementary service type of service selection model output and the main business type combination
Composite service；
From and more set predicting strategies of the main business type association in determine corresponding with objective cross business mesh
Predicting strategy is marked, and obtains the financial value of the objective cross business based on the target prediction strategy.
The second aspect of the embodiment of the present invention provides a kind of terminal device, and the terminal device includes memory, processing
Device and storage in the memory and the computer program that can run on the processor, described in the processor execution
Following steps are realized when computer program:
Multiple groups learning sample is obtained, and the multiple groups learning sample is handled by preset Processing Algorithm, is obtained
Service selection model, learning sample described in every group are made of main business sample, supplementary service sample and user's sample；
The service selection model is inputted using main business service type and the characteristic value of user object as prediction object information, and
It is the target of the user object by the supplementary service type of service selection model output and the main business type combination
Composite service；
From and more set predicting strategies of the main business type association in determine corresponding with objective cross business mesh
Predicting strategy is marked, and obtains the financial value of the objective cross business based on the target prediction strategy.
The third aspect of the embodiment of the present invention provides a kind of computer readable storage medium, the computer-readable storage
Media storage has computer program, and the computer program realizes following steps when being executed by processor:
Multiple groups learning sample is obtained, and the multiple groups learning sample is handled by preset Processing Algorithm, is obtained
Service selection model, learning sample described in every group are made of main business sample, supplementary service sample and user's sample；
The service selection model is inputted using main business service type and the characteristic value of user object as prediction object information, and
It is the target of the user object by the supplementary service type of service selection model output and the main business type combination
Composite service；
From and more set predicting strategies of the main business type association in determine corresponding with objective cross business mesh
Predicting strategy is marked, and obtains the financial value of the objective cross business based on the target prediction strategy.
Existing beneficial effect is the embodiment of the present invention compared with prior art:
The embodiment of the present invention is by obtaining multiple groups learning sample, wherein every group of learning sample is by main business sample, additional industry
Be engaged in sample and user's sample is constituted, and is then calculated by preset Processing Algorithm multiple groups learning sample, according to calculating
The result building service selection model arrived, it is then that main business service type and the characteristic value of user object is defeated as prediction object information
Enter to service selection model, the supplementary service type that service selection model is exported and main business type combination are user object
Objective cross business, finally by determining and objective cross industry in preset more set predicting strategies with main business type association
It is engaged in corresponding target prediction strategy, and based on target prediction policy calculation and obtains the financial value of objective cross business, the present invention
Embodiment can predict user object most probable and handle in the case where known main business service type and the characteristic value of user object
Objective cross business, and calculate financial value, improve the reliability of composite service earnings forecast.
Detailed description of the invention
It to describe the technical solutions in the embodiments of the present invention more clearly, below will be to embodiment or description of the prior art
Needed in attached drawing be briefly described, it should be apparent that, the accompanying drawings in the following description is only of the invention some
Embodiment for those of ordinary skill in the art without any creative labor, can also be according to these
Attached drawing obtains other attached drawings.
Fig. 1 is the implementation flow chart of the method for the automatic Prediction composite service income that the embodiment of the present invention one provides；
Fig. 2 is the implementation flow chart of the method for automatic Prediction composite service income provided by Embodiment 2 of the present invention；
Fig. 3 is the implementation flow chart of the method for the automatic Prediction composite service income that the embodiment of the present invention three provides；
Fig. 4 is the implementation flow chart of the method for the automatic Prediction composite service income that the embodiment of the present invention four provides；
Fig. 5 is the implementation flow chart of the method for the automatic Prediction composite service income that the embodiment of the present invention five provides；
Fig. 6 is the structural block diagram for the terminal device that the embodiment of the present invention six provides；
Fig. 7 is the schematic diagram for the terminal device that the embodiment of the present invention seven provides；
Fig. 8 is the schematic diagram for the service selection model that the embodiment of the present invention eight provides；
Fig. 9 is the schematic diagram for another service selection model that the embodiment of the present invention nine provides.
Specific embodiment
In being described below, for illustration and not for limitation, the tool of such as particular system structure, technology etc is proposed
Body details, to understand thoroughly the embodiment of the present invention.However, it will be clear to one skilled in the art that there is no these specific
The present invention also may be implemented in the other embodiments of details.In other situations, it omits to well-known system, device, electricity
The detailed description of road and method, in case unnecessary details interferes description of the invention.
In order to illustrate technical solutions according to the invention, the following is a description of specific embodiments.
Fig. 1 shows the implementation process of the method for automatic Prediction composite service income provided in an embodiment of the present invention, is described in detail
It is as follows:
In S101, multiple groups learning sample is obtained, and carry out to the multiple groups learning sample by preset Processing Algorithm
Processing, obtains service selection model, learning sample described in every group is by main business sample, supplementary service sample and user's sample structure
At.
Business refers to routine work to be treated, and in embodiments of the present invention, business is composite service, by main business
It is constituted with supplementary service, main business service type and supplementary service type indicate respectively the type of main business and supplementary service.In tradition
It is more that user is obtained by subjective judgement in main business under the premise of the main business that known users need to handle in method
On the basis of the supplementary service that may handle.And in embodiments of the present invention, multiple groups learning sample, one group of study sample are obtained first
It originally is the relevant information of in progress or completed composite service, every group of learning sample is by main business sample, supplementary service
Sample and user's sample are constituted, wherein it is noted that in the present embodiment main business service type and supplementary service type be towards
The appellation of user object, main business sample and supplementary service sample in every group of learning sample substantially respectively with main business service type
Identical with the meaning of supplementary service type, appellation difference, which is intended merely to facilitate, carries out area for learning sample and following user object
Point.
User's sample in every group of learning sample is the description to handling learning sample and corresponding to the user of composite service, user
Comprising multiple characteristic values, multiple known features can be pre- according to specific business correspondingly with multiple known features in sample
It first sets, for example multiple known features are gender, age and wage, characteristic value indicates the attribute of corresponding known features.It is optional
Ground, for ease of calculation, before being handled by preset Processing Algorithm multiple groups learning sample, to every group of learning sample
Multiple characteristic values of middle user's sample carry out section assignment operation again, assign corresponding calculated value to each characteristic value.Citing comes
It says, can be less than 50 years old and to be more than or equal to interval division 50 years old, if user's sample if some known features is the age
In the known features characteristic value less than 50 years old, then the calculated value of the known features in user's sample is assigned to 0；If user
The characteristic value of the known features is greater than or equal to 50 years old in sample, then assigns the calculated value of the known features in user's sample
It is 1, certainly, the division in section and the calculated value numerical value of imparting can be set according to practical application scene, i.e. section assignment again
It is not necessarily limited to above-mentioned example.
Optionally, multiple groups learning sample is obtained according to preset acquisition condition.In embodiments of the present invention, it can be preset
Acquisition condition goes out multiple groups learning sample according to acquisition conditional filtering from large batch of business sample, to composite service income
Carry out qualitative forecasting, wherein business sample is identical as the format of learning sample, refers to the combination industry of the embodiment of the present invention
All relevant informations of business, the quantity of ordinary business practice sample are more.Specific acquisition process is as follows: such as to the group to the city A
Conjunction business carries out earnings forecast, then it is " filtering out the learning sample that source place is the city A " that acquisition condition, which is arranged, from there are multiple
Learning sample is filtered out in the business sample in source place；Such as to be greater than quinquagenary combination to some known features such as age
Business carries out earnings forecast, then it is " filter out in user's sample age be greater than quinquagenary learning sample " that acquisition condition, which is arranged,.
Acquisition condition can be single condition, be also possible to the combination of more than two conditions, for example settable acquisition condition is " working
Source place is filtered out in business sample is greater than quinquagenary learning sample for the age in the city A and user's sample ", obtain the setting of condition
It is also not limited to above-mentioned example, can be determined according to practical application scene.In addition, in order to reduce obtain learning sample and
The pressure for carrying out subsequent processing, can also limit the quantity of learning sample, and the learning sample quantity got is made to be in pre-
If the order of magnitude, the order of magnitude can be freely arranged, for example the order of magnitude can be set to 1,000.
Optionally, multiple groups learning sample is obtained from service database.Service database can be operation composite service
Main body (such as mechanism or company) stores the database of composite service, usually, the business sample stored in service database
The composite service situation being related to is more, and the order of magnitude of business sample is high, therefore can be directly from the business sample of service database
Middle determining multiple groups learning sample.In addition, there is calculating error caused by deviation due to obtaining learning sample in order to prevent, therefore obtain employment
Random fashion is taken to select business sample as learning sample, or after the condition of acquisition determines, from business datum in business database
Library, which meets in the business sample of acquisition condition, takes random fashion to determine learning sample.
In embodiments of the present invention, after getting multiple groups learning sample, multiple groups are learnt by preset Processing Algorithm
Sample is handled.Processing mode is divided into two kinds, and first way is to make the main business service type sample in every group of learning sample
For a known features, with supplementary service sample and user's sample together as the input parameter of service selection model at
Reason；The second way is to classify first to multiple groups learning sample according to main business sample, by identical of main business sample
It practises sample and is classified as one kind, then the learning sample of every class is individually handled, i.e., in the second way, main business sample is only made
For the standard of multiple groups learning sample classification, it is not intended as input parameter.Treatment process is illustrated based on the second way, such as
Under:
Firstly, being calculated separately out more in user's sample in certain identical multiple groups learning sample of one kind main business sample
A known features descend such multiple conversion gains of multiple groups learning sample, let it be assumed, for the purpose of illustration, that user's sample is more
A known features include gender Sex, age Age and wage Salary, and supplementary service sample includes Additionaltype1With
Additionaltype2Two kinds, then every group of learning sample in such all includes the feature of the characteristic value of gender Sex, age Age
One of value, the characteristic value of wage Salary and above two supplementary service sample.To the multiple groups learning sample under such
It is handled to obtain service selection model, first has to calculate separately above three known features respectively for such lower multiple groups study
The conversion gain of sample, conversion gain indicate known features for the significance level of such lower multiple groups learning sample, conversion gain
Bigger, then significance level is higher.In embodiments of the present invention, gender Sex, age Age and wage Salary point are calculated first
Not for the conversion gain of all groups of learning samples under such, calculating process does not need to limit computation sequence, can be to gender
The computation sequence of Sex, age Age and wage Salary carry out any sequence.The mistake calculated with the conversion gain to gender Sex
Journey is illustrated.
Obtain the sample total number Number of the multiple groups learning sample under suchsample, and Number is setsample1With
Numbersample2Respectively representing supplementary service sample in the multiple groups learning sample under such is Additionaltype1、
Additionaltype2Number of samples, then calculate the first of the overall conversion gain of the multiple groups learning sample under such calculate it is public
Formula are as follows:
In above-mentioned first calculation formula, K is preset gain coefficient, and the overall conversion gain for making is convenient for system
Meter and calculating, such as K may be configured as 10.
Assuming that the calculated value of gender Sex includes two classes, respectively Sextype1And Sextype2, and assume
Numbertype1-sampleCalculated value for known features Sex in the multiple groups learning sample under such is Sextype1Number of samples,
And in Numbertype1-sampleIn corresponding learning sample, it is assumed that Numbertype1-sample1It is for supplementary service sample
Additionaltype1Number of samples, it is assumed that Numbertype1-sample2It is Additional for supplementary service sampletype2Sample
This number；Similarly, it is assumed that Numbertype2-sampleCalculated value for known features Sex in the multiple groups learning sample under such is
Sextype2Number of samples, and in Numbertype2-sampleIn corresponding learning sample, it is assumed that Numbertype2-sample1It is additional
Business sample is Additionaltype1Number of samples, it is assumed that Numbertype2-sample2It is for supplementary service sample
Additionaltype2Number of samples.Then calculate the calculated value Sex of gender Sextype1The second of corresponding conversion gain calculates
Formula are as follows:
Calculate the calculated value Sex of gender Sextype2The third calculation formula of corresponding conversion gain are as follows:
It is Sex by calculated valuetype1And Sextype2Regard the branch under gender Sex as, then according to the first calculation formula, second
Calculation formula and third calculation formula calculate the of conditional diffusion gain of the gender Sex relative to the multiple groups learning sample under such
Four calculation formula are as follows:
The conditional diffusion gain that 4th calculation formula obtains represents the multiple groups in the case where gender Sex is determined, under such
The uncertainty degree of learning sample, therefore can obtain calculating the diffusion of gender Sex according to overall conversion gain and conditional diffusion gain
5th calculation formula of gain:
GainExtended(Sex)=GainExtended(All)-GainExtended(All|Sex)
The conversion gain Gain of calculated gender SexExtended(Sex) determine gender Sex to the multiple groups under such
The significance level of learning sample.Similarly, the conversion gain Gain of age Age can be calculatedExtended(Age) and wage Salary
Conversion gain GainExtended(Salary).Determine GainExtended(Sex)、GainExtended(Age) and GainExtended
(Salary) the maximum conversion gain of numerical value in, and using the corresponding known features of the conversion gain as the of service selection model
One sample node (first sample node corresponds to all groups of learning samples under such), and the difference of the known features is counted
Calculation value divides downwards as branch.For example maximum conversion gain is GainExtended(Sex), then by two calculating of gender Sex
Value is used as Liang Ge branch, divides downwards to the multiple groups learning sample under such, one of sample node after division is gender
The corresponding calculated value of characteristic value be Sextype1Learning sample, another sample node after division is the characteristic value pair of gender
The calculated value answered is Sextype2Learning sample.After division, then the sample node after division is continued to calculate in addition to gender
Multiple known features except Sex will count multiple conversion gains of the corresponding learning sample of sample node after the division
It is worth the calculated value of the corresponding known features of maximum conversion gain as branch, the sample node after the division is divided once again
It splits, the continuous iteration above process, until all known features of user's sample have all been used as sample node, then industry
Business preference pattern building finishes.Fig. 8 is the service selection model schematic being calculated with the above-mentioned second way, is only made
For example, as shown in figure 8, age Age has calculated value Agetype1And Agetype2, wage Salary has calculated value Salarytype1With
Salarytype2, gender Sex is the sample node of the first order in service selection model, and age Age and wage Salary are the second level
Sample node, most junior is different the supplementary service sample that calculated value obtains, including Additionaltype1With
Additionaltype2.In the sample node of age Age, it is corresponding be gender Sex calculated value be Sextype1Multiple groups
Practise sample.It is noted that there is the learning sample of N number of classification, then structure if there is N number of different main business sample
There is also N number of for the service selection model built.
In addition, Fig. 9 is another service selection model schematic being calculated with above-mentioned first way, only make
For example, as shown in figure 9, main business sample includes two two types of sample one and sample, gender Sex has calculated value Sextype1With
Sextype2, age Age has calculated value Agetype1And Agetype2, wage Salary has calculated value Salarytype1With
Salarytype2, supplementary service sample includes Additionaltype1、Additionaltype2And Additionaltype3Three types
Type.By Fig. 9, main business sample is that the supplementary service sample of the corresponding all groups of learning samples of sample two is all
Additionaltype3, then after obtaining the branched structure that main business sample is sample two, it is no longer based on the branched structure and divides downwards
It splits.
In S102, main business service type and the characteristic value of user object are inputted into the business as prediction object information and selected
Model is selected, and is the user couple by the supplementary service type of service selection model output and the main business type combination
The objective cross business of elephant.
After service selection model generates, the characteristic value of main business service type and user object is obtained, and by main business
Type and the characteristic value of user object are as the prediction object information input service selection model, wherein the spy of user object
Value indicative is consistent with the format of user's sample, such as inclusive another characteristic value in user's sample, the characteristic value at age and wage
Characteristic value, then the characteristic value of user object is also required to the characteristic value of inclusive another characteristic value, the characteristic value at age and wage.Root
According to the difference of service selection model generating mode, after prediction object information is input to service selection model, the mode of calculating
There are differences, are by the main business sample in learning sample if service selection model is that above-mentioned first way is taken to generate
This conduct inputs parameter, then in this step, directly be calculated to prediction object information by service selection model attached
Add type of service；If service selection model is that the above-mentioned second way is taken to generate, first from multiple service selection models
Service selection model corresponding with the main business service type in prediction object information is selected, then the characteristic value of user object is input to
The service selection model, to obtain the supplementary service type of service selection model output.Getting service selection model
It is objective cross business by supplementary service type and main business type combination after the supplementary service type of output.
In S103, the determining and objective cross business from more set predicting strategies with the main business type association
Corresponding target prediction strategy, and obtain based on the target prediction strategy financial value of the objective cross business.
For a main business service type, may from different supplementary service type combinations, to be formed different
Composite service, and the income calculation mode of different composite service is also likely to be present difference.Therefore it obtains and main business type association
Multiple composite service more set predicting strategies, wherein predicting strategy can be for formula or calculating logic etc..According to step S102
The objective cross business of middle determination determines target prediction strategy corresponding with objective cross business from more set predicting strategies, and
Objective cross business is calculated based on target prediction strategy, obtains financial value.It is noted that due to objective cross industry
Business may be only related with the type of business, without including specific service parameter, therefore before calculating financial value, acquisition and target
The corresponding service parameter of composite service.Objective cross business is calculated, is to be joined based on target prediction strategy to business actually
The process that number is calculated.It is a kind of citing of combination policy business with objective cross business, which includes main danger
Type and attached dangerous type, and the corresponding target prediction strategy of combination policy business be using attached dangerous premium as intermediate premium, then
Using the sum of intermediate premium and main dangerous premium as main dangerous protection amount, calculated main dangerous protection amount is combination policy business
Financial value.Then after determining target prediction strategy, attached dangerous premium and main dangerous premium are obtained as service parameter.
In embodiments of the present invention, the presence of objective cross business is uniqueness, and there is no inclusion relations, such as main business
Service type is A, and the objective cross business that supplementary service type is B and main business service type are A, and supplementary service type is the mesh of B and C
Inclusion relation is marked between composite service and be not present, is two individual composite service.
Wherein, more set predicting strategies can be stored in EXCEL table or database, and script file is set, execute foot
During this document, if detect objective cross business determine after, automatically from EXCEL table or database obtain and target
The corresponding target prediction strategy of composite service, and according to the financial value of target prediction policy calculation objective cross business, it improves
The degree of automation of income calculation.
By embodiment illustrated in fig. 1 it is found that in embodiments of the present invention, passing through acquisition multiple groups learning sample, every group of study
Sample is made of main business sample, supplementary service sample and user's sample, is counted by Processing Algorithm to multiple groups learning sample
Calculation obtains service selection model, then inputs the industry using main business service type and the characteristic value of user object as prediction object information
The supplementary service type of service selection model output and main business type combination are the target of user object by business preference pattern
Composite service, finally from and more set predicting strategies of main business type association in determine that corresponding with objective cross business target is pre-
Strategy is surveyed, and obtains the financial value of objective cross business based on target prediction strategy, the embodiment of the present invention is selected by building business
Model is selected, the accuracy of objective cross traffic forecast and financial value prediction is improved.
Shown in Fig. 2, be on the basis of the embodiment of the present invention one, to by Processing Algorithm to multiple groups learning sample at
Obtained a kind of implementation method after the step of reason is refined.The embodiment of the invention provides automatic Prediction composite service incomes
The implementation flow chart of method, as shown in Fig. 2, the method for the automatic Prediction composite service income may comprise steps of:
In S201, preset sample value range corresponding with user's sample is obtained.
Learning sample generates and the process of storage is likely to occur mistake or program error, leads to the multiple groups got
It practises in sample there may be exceptional sample, since exceptional sample is usually the spy of some or certain several known features of user's sample
There is exception in value indicative, for example the characteristic value at age is negative, therefore in embodiments of the present invention, obtain preset and user's sample pair
The sample value range answered, sample value range include multiple feature value models corresponding with known features multiple in user's sample
It encloses.Wherein, feature value range can be formulated according to limited case of the composite service to known features, such as certain insurance limitation
The age of insurer is ten years old to 70 years old, then age corresponding feature value range is ten to 70.
In S202, user's sample therein is filtered out from the multiple groups learning sample and is in the sample value range
Learning sample, and the learning sample filtered out is exported to learning sample collection.
After getting the sample value range of user's sample, multiple groups learning sample is examined based on sample value range
It surveys, judges whether user's sample of every group of learning sample is located in sample value range, specifically judge to use under every group of learning sample
Whether the characteristic value of multiple known features of family sample is located in sample value range in corresponding feature value range.If certain group
User's sample of learning sample is located in sample value range, then exports this group of learning sample to learning sample collection.
Optionally, it if the characteristic value for detecting some known features of user's sample in certain group learning sample is sky, issues
Warning note leaves out this group of learning sample.If the characteristic value of some known features of user's sample is sky in certain group learning sample
Value then may be loss of data or corrupt data, then can issue the user with warning note, prompts user in this group of learning sample
The characteristic value of the known features of user's sample carries out numerical value supplement or value revision, or directly leaves out this group of learning sample.
When user in this group of learning sample user's sample values supplement or value revision after, can by this group of learning sample again on
It passes, if detecting, the characteristic value of all known features of user's sample in this group of learning sample is in corresponding feature value model
After enclosing, this group of learning sample is exported to learning sample collection automatically.
Learning sample is being exported to learning sample collection, the repeated sample that can also concentrate to learning sample detects,
As shown in figure 3, the method for the automatic Prediction composite service income may comprise steps of:
In S301, detects the learning sample and concentrate with the presence or absence of repeated sample.
In practical business, it is possible that identical learning sample, the identical main business sample referred in learning sample,
Supplementary service sample and user's sample are all the same, and identical reason occur may be to have carried out repeating to deposit to learning sample in advance
Storage, to prevent loss of data；It is also likely to be simple memory error etc..Therefore in embodiments of the present invention, learning sample is concentrated
Multiple groups learning sample detected, detection learning sample, which is concentrated, actually whether there is repeated sample, and repeated sample refers to main business
Business sample, supplementary service sample and the identical learning sample more than two of user's sample standard deviation.
In S302, if detecting the repeated sample, only retain the repeated sample in learning sample concentration
In one group described in learning sample.
Since service selection model is to be calculated according to the number of learning sample, therefore repeated sample can be to follow-up business
The generation of preference pattern impacts, and makes the reduction of the laminating degree of service selection model and actual conditions.Therefore implement in the present invention
In example, if detecting repeated sample, the one group of learning sample only retained in repeated sample is concentrated in learning sample, is effectively mentioned
The accuracy of service selection model generation is risen.
In S203, the learning sample collection is handled by the Processing Algorithm.
To multiple groups learning sample screening after, by preset Processing Algorithm to the learning sample collection of generation at
Reason, to construct service selection model.
It is preset corresponding with user's sample by obtaining by embodiment illustrated in fig. 2 it is found that in embodiments of the present invention
Sample value range filters out the learning sample that user's sample therein is in sample value range from multiple groups learning sample,
And export the learning sample filtered out to learning sample collection, learning sample collection is handled finally by Processing Algorithm, is had
Effect eliminates the exceptional sample in multiple groups learning sample, and it is unfavorable to prevent exceptional sample to cause the generation of service selection model
It influences.
It is on the basis of the embodiment of the present invention one, if desired the objective cross business of calculating income is more shown in Fig. 4
It is a, then it is the objective cross industry of user object to the supplementary service type and main business type combination that export service selection model
A kind of implementation method that business obtains after being refined.The embodiment of the invention provides the methods of automatic Prediction composite service income
Implementation flow chart, as shown in figure 4, the method for the automatic Prediction composite service income may comprise steps of:
In S401, the supplementary service type that the service selection model exports is added to supplementary service collection.
In practical application scene, exists after main business service type and the characteristic value of user object determine, need to obtain more
A objective cross business, and calculate the demand of multiple financial values.Therefore in embodiments of the present invention, first by service selection model
The supplementary service type exported based on main business service type and the characteristic value of user object is added to supplementary service collection.
In S402, traced back in the service selection model based on the supplementary service type, obtain with it is described
The sample node of the highest user's sample of the supplementary service type degree of correlation, the sample node and user's sample are
Know feature correlation.
After obtaining the supplementary service type of service selection model output, since the supplementary service type is in service selection model
In be determined by a certain branched structure, therefore the branched structure according to belonging to the supplementary service type in service selection model into
Row traces back, and finds the sample node with the immediate user's sample of the supplementary service type, the sample node and user's sample
In known features it is corresponding, the degree of correlation of the sample node and the supplementary service type also highest.As shown in figure 9, if main business
Type is sample one, and the corresponding calculated value of the characteristic value of user object is Sextype1And Agetype1, then the supplementary service class that obtains
Type is Additionaltype1, traced back based on the supplementary service type, obtain with the supplementary service type degree of correlation highest
Sample node be age Age.
In S403, the subactivity type in the service selection model is obtained, and the subactivity type is added
The supplementary service collection is added to, the subactivity type is the sample node subordinate except the service selection model exports
The supplementary service type outside supplementary service type.
In embodiments of the present invention, multiple objective cross business are if desired generated, then after main business service type determines, preferentially
The lesser characteristic value of influence degree in the characteristic value of user object is changed by consideration, and by the spy of the user object after change
Value indicative corresponding supplementary service type in service selection model is added to supplementary service collection.Specifically, in service selection mould
In type, there may be one or more branches to tie by subordinate for each sample node (each known features of user's sample)
Structure, therefore obtain in service selection model, and the highest sample node subordinate of supplementary service type association degree, except service selection
Other supplementary service types except the supplementary service type of model output, distinguish for convenience, are named as secondary industry
Service type, and subactivity type is added to supplementary service collection.If the highest sample node subordinate of the degree of correlation is simultaneously not present
Other branched structures, that is, only exist the supplementary service type of service selection model output, or subactivity type is added to
After supplementary service collection, do not meet the preset the upper limit of the number of supplementary service collection yet, then the highest sample node of the degree of correlation is based on, in industry
Continue to trace back in business preference pattern, determine the more supervisory sample node of the highest sample node of the degree of correlation, and obtain with more
The relevant subactivity type of supervisory sample node.
In S404, the supplementary service type and the subactivity type that the supplementary service is concentrated respectively with
The main business type combination is multiple objective cross business.
(judgment criteria that generation finishes may be to reach the upper limit of the number of supplementary service collection after supplementary service collection generates
Or reach other preset conditions), by supplementary service concentrate supplementary service type and subactivity type respectively with main business class
Type is combined, and obtains multiple objective cross business.
By embodiment illustrated in fig. 4 it is found that in embodiments of the present invention, passing through the additional industry for exporting service selection model
Service type is added to supplementary service collection, and is traced back in service selection model based on supplementary service type, obtains and adds
The sample node of the highest user's sample of the type of service degree of correlation, the sample node is related to the known features of user's sample, so
The subactivity type in addition to supplementary service type for obtaining sample node subordinate in service selection model afterwards, finally will be attached
The supplementary service type and subactivity type for adding service set are respectively multiple objective cross business with main business type combination,
Objective cross business suitable for demand is multiple application scenarios, improves and needs to carry out earnings forecast to multiple composite service
In the case of, the reliability of multiple objective cross business of generation.
Shown in Fig. 5, be on the basis of the embodiment of the present invention four, if supplementary service collection there are the upper limit of the number, and secondary industry
Service type is there are multiple, then to the subactivity type obtained in service selection model, and subactivity type is added to attached
A kind of implementation method obtained after adding services sets to be refined.The embodiment of the invention provides automatic Prediction composite service incomes
The implementation flow chart of method, as shown in figure 5, the method for the automatic Prediction composite service income may comprise steps of:
In S501, multiple yield values of the sample node obtained by the Processing Algorithm are obtained, wherein described
Multiple yield values and multiple subactivity types are related one by one.
After the determining highest sample node of the supplementary service type degree of correlation with the output of service selection model, due in business
In preference pattern, which usually has multiple branched structures, and the supplementary service type under each branched structure is different, therefore
Multiple yield values of the sample node can be calculated based on branched structure, which is substantially that the sample node is corresponding known
Multiple calculated values of feature are respectively relative to the conditional diffusion gain of multiple groups learning sample.Specifically, it is calculated with the above-mentioned the 4th
Formula citing, calculated GainExtended(ALL | Sex) is the conditional diffusion gain of gender Sex, then under gender Sex
Two class calculated value Sextype1And Sextype2, corresponding conditional diffusion gain are as follows:
Sextype1And Sextype2Corresponding two conditional diffusion gain Gs ainExtended(ALL|Sextype1) and
GainExtended(ALL|Sextype2) be known features gender Sex two yield values.Similarly with above-mentioned calculation method, exist
After the determining highest sample node (known features) of the supplementary service type degree of correlation with the output of service selection model, obtain with
Know the relevant multiple calculated values of feature multiple yield values correspondingly.
In S502, multiple subactivity types are ranked up according to the numerical values recited of the multiple yield value,
Generate subactivity sequence.
Since conditional diffusion gain represents after certain condition determines, other conditions to the influence degree of multiple groups learning sample,
Therefore conditional diffusion gain is bigger, then the corresponding condition of conditional diffusion gain is smaller to the influence degree of multiple groups learning sample.Therefore it presses
It is ranked up according to numerical values recited multiple subactivity types corresponding to multiple yield values of multiple yield values, sort method is preferred
It is from small to large, wherein each yield value corresponds to a branched structure in service selection model, i.e., a corresponding secondary industry
Service type.After the completion of sequence, subactivity sequence is generated.
In S503, the subactivity type is successively chosen from front to back from the subactivity sequence, and add
To the supplementary service collection, the upper limit of the number until reaching the supplementary service collection.
Subactivity sequence generation after, from subactivity sequence from front to back (i.e. according to sequence from small to large) according to
Secondary selection subactivity type, and it is added sequentially to supplementary service collection, when the quantity that supplementary service is concentrated reaches the upper limit of the number
When, then stop adding.If not reached yet after all subactivity types in subactivity sequence are all added to supplementary service collection
To the upper limit of the number of supplementary service collection, then in the immediate degree of correlation highest of supplementary service type exported with service selection model
Sample nodal basis on, continuation is traced back in service selection model, obtains more supervisory sample node, and calculate upper
Multiple yield values of the sample node of grade generate subactivity sequence, and choose secondary from front to back from subactivity sequence
Type of service is added to supplementary service collection, the continuous iteration above process, until reaching the upper limit of the number of supplementary service collection.It can
Selection of land, in order to avoid adding duplicate subactivity type, in the subactivity type for obtaining sample node subordinate or from secondary
When choosing subactivity type in traffic sequence, judge that supplementary service is concentrated with the presence or absence of identical subactivity type, if depositing
Then omitting the subactivity type；If it does not exist, then normal execute obtains subactivity type or selection subactivity type
Operation.
By embodiment illustrated in fig. 5 it is found that in embodiments of the present invention, by obtaining the sample obtained by Processing Algorithm
Multiple yield values of node, multiple yield values and multiple subactivity types are related one by one, and according to the numerical value of multiple yield values
Size is ranked up multiple subactivity types, generates subactivity sequence, from subactivity sequence from front to back successively
Subactivity type is chosen, and subactivity type is added to supplementary service collection, the number until reaching supplementary service collection
Until measuring the upper limit, on the basis of sample node, that is, known features determine, user's most probable is predicted by calculating multiple yield values
The subactivity type of selection improves the reliability for generating multiple objective cross business.
It should be understood that the size of the serial number of each step is not meant that the order of the execution order in above-described embodiment, each process
Execution sequence should be determined by its function and internal logic, the implementation process without coping with the embodiment of the present invention constitutes any limit
It is fixed.
Fig. 6 shows the structural block diagram of terminal device provided in an embodiment of the present invention, each unit which includes
For executing each step in the corresponding embodiment of Fig. 1.It is retouched referring specifically to Fig. 1 is related in embodiment corresponding to Fig. 1
It states.For ease of description, only the parts related to this embodiment are shown.
Referring to Fig. 6, the terminal device includes:
Processing unit 61 learns sample to the multiple groups for obtaining multiple groups learning sample, and by preset Processing Algorithm
This is handled, and obtains service selection model, learning sample described in every group is by main business sample, supplementary service sample and user's sample
This composition；
Assembled unit 62, for using the characteristic value of main business service type and user object as described in prediction object information input
Service selection model, and be described by the supplementary service type of service selection model output and the main business type combination
The objective cross business of user object；
Acquiring unit 63, for the determining and target group from more set predicting strategies with the main business type association
The corresponding target prediction strategy of conjunction business, and obtain based on the target prediction strategy financial value of the objective cross business.
Optionally, the processing unit 61, comprising:
Range acquiring unit, for obtaining preset sample value range corresponding with user's sample；
Output unit is in the sample value for filtering out user's sample therein from the multiple groups learning sample
The learning sample of range, and the learning sample filtered out is exported to learning sample collection；
Subelement is handled, for handling by the Processing Algorithm the learning sample collection.
Optionally, the output unit, further includes:
Detection unit is concentrated for detecting the learning sample with the presence or absence of repeated sample；
Stick unit, if only retaining the repetition in learning sample concentration for detecting the repeated sample
Learning sample described in one group in sample.
Optionally, the assembled unit 62, comprising:
First adding unit, the supplementary service type for exporting the service selection model are added to supplementary service
Collection；
Trace back unit, for being traced back in the service selection model based on the supplementary service type, obtain with
The sample node of the highest user's sample of the supplementary service type degree of correlation, the sample node and user's sample
Known features it is related；
Second adding unit, for obtaining the subactivity type in the service selection model, and by the secondary industry
Service type is added to the supplementary service collection, and the subactivity type is the sample node subordinate except the service selection
Supplementary service type outside the supplementary service type of model output；
Combine subelement, the supplementary service type and the subactivity type for concentrating the supplementary service
It is respectively multiple objective cross business with the main business type combination.
Optionally, if there are the upper limit of the number for supplementary service collection, and there are multiple, then second additions for subactivity type
Unit, comprising:
Gain acquiring unit, for obtaining multiple yield values of the sample node obtained by the Processing Algorithm,
Wherein, the multiple yield value and multiple subactivity types are related one by one；
Sequencing unit arranges multiple subactivity types for the numerical values recited according to the multiple yield value
Sequence generates subactivity sequence；
Business selection unit, for successively choosing the subactivity class from front to back from the subactivity sequence
Type, and it is added to the supplementary service collection, the upper limit of the number until reaching the supplementary service collection.
Therefore, terminal device provided in an embodiment of the present invention is realized by building service selection model to objective cross
Business and prediction to financial value, improve forecasting reliability.
Fig. 7 is the schematic diagram of terminal device provided in an embodiment of the present invention.As shown in fig. 7, the terminal device 7 of the embodiment
Include: processor 70, memory 71 and is stored in the calculating that can be run in the memory 71 and on the processor 70
Machine program 72.The processor 70 realizes above-mentioned each automatic Prediction composite service income when executing the computer program 72
Step in embodiment of the method, such as step S101 to S103 shown in FIG. 1.Alternatively, the processor 70 executes the calculating
The function of each unit in above-mentioned each terminal device embodiment, such as the function of unit 61 to 63 shown in Fig. 6 are realized when machine program 72.
Illustratively, the computer program 72 can be divided into one or more units, one or more of
Unit is stored in the memory 71, and is executed by the processor 70, to complete the present invention.One or more of lists
Member can be the series of computation machine program instruction section that can complete specific function, and the instruction segment is for describing the computer journey
Implementation procedure of the sequence 72 in the terminal device 7.For example, the computer program 72 can be divided into processing unit, group
It closes unit and acquiring unit, each unit concrete function is as follows:
Processing unit, for obtaining multiple groups learning sample, and by preset Processing Algorithm to the multiple groups learning sample
It is handled, obtains service selection model, learning sample described in every group is by main business sample, supplementary service sample and user's sample
It constitutes；
Assembled unit, for inputting the industry for the characteristic value of main business service type and user object as prediction object information
Business preference pattern, and be the use by the supplementary service type of service selection model output and the main business type combination
The objective cross business of family object；
Acquiring unit, for the determining and objective cross from more set predicting strategies with the main business type association
The corresponding target prediction strategy of business, and obtain based on the target prediction strategy financial value of the objective cross business.
The terminal device 7 can be the calculating such as desktop PC, notebook, palm PC and cloud server and set
It is standby.The terminal device 7 may include, but be not limited only to, processor 70, memory 71.It will be understood by those skilled in the art that figure
7 be only the example of terminal device 7, does not constitute the restriction to terminal device 7, may include than illustrating more or fewer portions
Part perhaps combines certain components or different components, such as the terminal device 7 can also include input-output equipment, net
Network access device, bus etc..
Alleged processor 70 can be central processing unit (Central Processing Unit, CPU), can also be
Other general processors, digital signal processor (Digital Signal Processor, DSP), specific integrated circuit
(Application Specific Integrated Circuit, ASIC), ready-made programmable gate array (Field-
Programmable Gate Array, FPGA) either other programmable logic device, discrete gate or transistor logic,
Discrete hardware components etc..General processor can be microprocessor or the processor is also possible to any conventional processor
Deng.
The memory 71 can be the internal storage unit of the terminal device 7, such as the hard disk or interior of terminal device 7
It deposits.The memory 71 is also possible to the External memory equipment of the terminal device 7, such as be equipped on the terminal device 7
Plug-in type hard disk, intelligent memory card (Smart Media Card, SMC), secure digital (Secure Digital, SD) card dodge
Deposit card (Flash Card) etc..Further, the memory 71 can also both include the storage inside list of the terminal device 7
Member also includes External memory equipment.The memory 71 is for storing needed for the computer program and the terminal device 7
Other programs and data.The memory 71 can be also used for temporarily storing the data that has exported or will export.
It is apparent to those skilled in the art that for convenience of description and succinctly, only with above-mentioned each function
Can unit division progress for example, in practical application, can according to need and by above-mentioned function distribution by different functions
Unit is completed, i.e., the internal structure of the terminal device is divided into different functional units, to complete whole described above
Or partial function.Each functional unit in embodiment can integrate in one processing unit, be also possible to each unit list
It is solely physically present, can also be integrated in one unit with two or more units, above-mentioned integrated unit can both use
Formal implementation of hardware can also be realized in the form of software functional units.In addition, the specific name of each functional unit also only
It is the protection scope that is not intended to limit this application for the ease of mutually distinguishing.The specific work process of unit in above system,
It can refer to corresponding processes in the foregoing method embodiment, details are not described herein.
In the above-described embodiments, it all emphasizes particularly on different fields to the description of each embodiment, is not described in detail or remembers in some embodiment
The part of load may refer to the associated description of other embodiments.
Those of ordinary skill in the art may be aware that list described in conjunction with the examples disclosed in the embodiments of the present disclosure
Member and algorithm steps can be realized with the combination of electronic hardware or computer software and electronic hardware.These functions are actually
It is implemented in hardware or software, the specific application and design constraint depending on technical solution.Professional technician
Each specific application can be used different methods to achieve the described function, but this realization is it is not considered that exceed
The scope of the present invention.
In embodiment provided by the present invention, it should be understood that disclosed device/terminal device and method, it can be with
It realizes by another way.For example, device described above/terminal device embodiment is only schematical, for example, institute
The division of unit is stated, only a kind of logical function partition, there may be another division manner in actual implementation, such as multiple lists
Member or component can be combined or can be integrated into another system, or some features can be ignored or not executed.Another point,
Shown or discussed mutual coupling or direct-coupling or communication connection can be through some interfaces, device or unit
INDIRECT COUPLING or communication connection, can be electrical property, mechanical or other forms.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme
's.
It, can also be in addition, the functional units in various embodiments of the present invention may be integrated into one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also realize in the form of software functional units.
If the integrated unit is realized in the form of SFU software functional unit and sells or use as independent product
When, it can store in a computer readable storage medium.Based on this understanding, the present invention realizes above-described embodiment side
All or part of the process in method can also instruct relevant hardware to complete, the computer by computer program
Program can be stored in a computer readable storage medium, and the computer program is when being executed by processor, it can be achieved that above-mentioned each
The step of a embodiment of the method.Wherein, the computer program includes computer program code, and the computer program code can
Think source code form, object identification code form, executable file or certain intermediate forms etc..The computer-readable medium can be with
It include: any entity or device, recording medium, USB flash disk, mobile hard disk, magnetic disk, light that can carry the computer program code
Disk, computer storage, read-only memory (Read-Only Memory, ROM), random access memory (Random Access
Memory, RAM), electric carrier signal, telecommunication signal and software distribution medium etc..It should be noted that described computer-readable
The content that medium includes can carry out increase and decrease appropriate according to the requirement made laws in jurisdiction with patent practice, such as at certain
A little jurisdictions do not include electric carrier signal and telecommunication signal according to legislation and patent practice, computer-readable medium.
Embodiment described above is merely illustrative of the technical solution of the present invention, rather than its limitations；Although referring to aforementioned reality
Applying example, invention is explained in detail, those skilled in the art should understand that: it still can be to aforementioned each
Technical solution documented by embodiment is modified or equivalent replacement of some of the technical features；And these are modified
Or replacement, the spirit and scope for technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution should all
It is included within protection scope of the present invention.
Claims (10)
1. a kind of method of automatic Prediction composite service income characterized by comprising
Multiple groups learning sample is obtained, and the multiple groups learning sample is handled by preset Processing Algorithm, obtains business
Preference pattern, learning sample described in every group are made of main business sample, supplementary service sample and user's sample；
The service selection model is inputted using main business service type and the characteristic value of user object as prediction object information, and by institute
The supplementary service type and the main business type combination of stating the output of service selection model are the objective cross of the user object
Business；
From and more set predicting strategies of the main business type association in determine that corresponding with objective cross business target is pre-
Strategy is surveyed, and obtains the financial value of the objective cross business based on the target prediction strategy.
2. the method as described in claim 1, which is characterized in that described to learn sample to the multiple groups by preset Processing Algorithm
This is handled, comprising:
Obtain preset sample value range corresponding with user's sample；
The learning sample that user's sample therein is in the sample value range is filtered out from the multiple groups learning sample, and
The learning sample filtered out is exported to learning sample collection；
The learning sample collection is handled by the Processing Algorithm.
3. method according to claim 2, which is characterized in that the learning sample that will be filtered out is exported to study sample
After this collection, further includes:
The learning sample is detected to concentrate with the presence or absence of repeated sample；
If detecting the repeated sample, described in one group in the repeated sample is only retained in learning sample concentration
Practise sample.
4. the method as described in claim 1, which is characterized in that the supplementary service class for exporting the service selection model
Type and the main business type combination are the objective cross business of the user object, comprising:
The supplementary service type that the service selection model exports is added to supplementary service collection；
It is traced back, is obtained and the supplementary service type phase in the service selection model based on the supplementary service type
The sample node of the highest user's sample of Guan Du, the sample node are related to the known features of user's sample；
The subactivity type in the service selection model is obtained, and the subactivity type is added to the additional industry
Business collection, the subactivity type are the supplementary services of the sample node subordinate exported except the service selection model
Supplementary service type outside type；
The supplementary service type and the subactivity type that the supplementary service is concentrated respectively with the main business class
Type group is combined into multiple objective cross business.
5. method as claimed in claim 4, which is characterized in that if the supplementary service collection there are the upper limit of the number, and described time
Grade type of service is there are multiple, then the subactivity type obtained in the service selection model, and by the secondary industry
Service type is added to the supplementary service collection, comprising:
Obtain the multiple yield values of the sample node obtained by the Processing Algorithm, wherein the multiple yield value and
Multiple subactivity types are related one by one；
Multiple subactivity types are ranked up according to the numerical values recited of the multiple yield value, generate subactivity sequence
Column；
It successively chooses the subactivity type from front to back from the subactivity sequence, and is added to the supplementary service
Collection, the upper limit of the number until reaching the supplementary service collection.
6. a kind of terminal device, which is characterized in that the terminal device includes memory, processor and is stored in the storage
In device and the computer program that can run on the processor, the processor are realized as follows when executing the computer program
Step:
Multiple groups learning sample is obtained, and the multiple groups learning sample is handled by preset Processing Algorithm, obtains business
Preference pattern, learning sample described in every group are made of main business sample, supplementary service sample and user's sample；
The service selection model is inputted using main business service type and the characteristic value of user object as prediction object information, and by institute
The supplementary service type and the main business type combination of stating the output of service selection model are the objective cross of the user object
Business；
From and more set predicting strategies of the main business type association in determine that corresponding with objective cross business target is pre-
Strategy is surveyed, and obtains the financial value of the objective cross business based on the target prediction strategy.
7. terminal device as claimed in claim 6, which is characterized in that it is described by preset Processing Algorithm to the multiple groups
Sample is practised to be handled, comprising:
Obtain preset sample value range corresponding with user's sample；
The learning sample that user's sample therein is in the sample value range is filtered out from the multiple groups learning sample, and
The learning sample filtered out is exported to learning sample collection；
The learning sample collection is handled by the Processing Algorithm.
8. terminal device as claimed in claim 7, which is characterized in that the learning sample that will be filtered out is exported to
After habit sample set, further includes:
The learning sample is detected to concentrate with the presence or absence of repeated sample；
If detecting the repeated sample, described in one group in the repeated sample is only retained in learning sample concentration
Practise sample.
9. terminal device as claimed in claim 6, which is characterized in that the additional industry for exporting the service selection model
Service type and the main business type combination are the objective cross business of the user object, comprising:
The supplementary service type that the service selection model exports is added to supplementary service collection；
It is traced back, is obtained and the supplementary service type phase in the service selection model based on the supplementary service type
The sample node of the highest user's sample of Guan Du, the sample node are related to the known features of user's sample；
The subactivity type in the service selection model is obtained, and the subactivity type is added to the additional industry
Business collection, the subactivity type are the supplementary services of the sample node subordinate exported except the service selection model
Supplementary service type outside type；
The supplementary service type and the subactivity type that the supplementary service is concentrated respectively with the main business class
Type group is combined into multiple objective cross business.
10. a kind of computer readable storage medium, the computer-readable recording medium storage has computer program, and feature exists
In when the computer program is executed by processor the step of any one of such as claim 1 to 5 of realization the method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810554475.3A CN108961071B (en) | 2018-06-01 | 2018-06-01 | Method for automatically predicting combined service income and terminal equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810554475.3A CN108961071B (en) | 2018-06-01 | 2018-06-01 | Method for automatically predicting combined service income and terminal equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108961071A true CN108961071A (en) | 2018-12-07 |
| CN108961071B CN108961071B (en) | 2023-07-21 |
Family
ID=64493333
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810554475.3A Active CN108961071B (en) | 2018-06-01 | 2018-06-01 | Method for automatically predicting combined service income and terminal equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108961071B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110020728A (en) * | 2018-12-25 | 2019-07-16 | 阿里巴巴集团控股有限公司 | Business model intensified learning method and device |
| CN113965563A (en) * | 2020-12-18 | 2022-01-21 | 深圳平安智汇企业信息管理有限公司 | Model-based business processing method and device and server |
Citations (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2006134183A (en) * | 2004-11-08 | 2006-05-25 | Nippon Telegr & Teleph Corp <Ntt> | Information classification method, system and program, and storage medium with program stored |
| US20060190310A1 (en) * | 2005-02-24 | 2006-08-24 | Yasu Technologies Pvt. Ltd. | System and method for designing effective business policies via business rules analysis |
| US20100082421A1 (en) * | 2008-09-30 | 2010-04-01 | Yahoo! Inc. | Click through rate prediction system and method |
| US20140310053A1 (en) * | 2013-04-10 | 2014-10-16 | Xerox Corporation | Method and systems for providing business process suggestions and recommendations utilizing a business process modeler |
| CN105868298A (en) * | 2016-03-23 | 2016-08-17 | 华南理工大学 | Mobile phone game recommendation method based on binary decision tree |
| US20170004513A1 (en) * | 2015-07-01 | 2017-01-05 | Rama Krishna Vadakattu | Subscription churn prediction |
| CN106817248A (en) * | 2016-12-19 | 2017-06-09 | 西安电子科技大学 | A kind of APT attack detection methods |
| CN107122390A (en) * | 2017-03-04 | 2017-09-01 | 华数传媒网络有限公司 | Recommendation system building method based on groups of users |
| WO2017186048A1 (en) * | 2016-04-27 | 2017-11-02 | 第四范式（北京）技术有限公司 | Method and device for presenting prediction model, and method and device for adjusting prediction model |
| WO2017219548A1 (en) * | 2016-06-20 | 2017-12-28 | 乐视控股（北京）有限公司 | Method and device for predicting user attributes |
| CN107679686A (en) * | 2017-08-28 | 2018-02-09 | 阿里巴巴集团控股有限公司 | A kind of business performs method and device |
| CN107679994A (en) * | 2017-07-26 | 2018-02-09 | 平安科技（深圳）有限公司 | Core protects creation method, device, computer equipment and the storage medium of decision tree |
| CN107689008A (en) * | 2017-06-09 | 2018-02-13 | 平安科技（深圳）有限公司 | A kind of user insures the method and device of behavior prediction |
| CN107730389A (en) * | 2017-09-30 | 2018-02-23 | 平安科技（深圳）有限公司 | Electronic installation, insurance products recommend method and computer-readable recording medium |
| CN107818344A (en) * | 2017-10-31 | 2018-03-20 | 上海壹账通金融科技有限公司 | The method and system that user behavior is classified and predicted |
| US9928552B1 (en) * | 2009-03-13 | 2018-03-27 | New York Life Insurance Company | Methods and systems for insurance investment product decision modeling |
| CN107944990A (en) * | 2017-12-29 | 2018-04-20 | 山大地纬软件股份有限公司 | A kind of integral counter-employee device and method of the precision push based on machine learning |
| CN107944629A (en) * | 2017-11-30 | 2018-04-20 | 北京邮电大学 | A kind of recommendation method and device based on heterogeneous information network representation |
-
2018
- 2018-06-01 CN CN201810554475.3A patent/CN108961071B/en active Active
Patent Citations (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2006134183A (en) * | 2004-11-08 | 2006-05-25 | Nippon Telegr & Teleph Corp <Ntt> | Information classification method, system and program, and storage medium with program stored |
| US20060190310A1 (en) * | 2005-02-24 | 2006-08-24 | Yasu Technologies Pvt. Ltd. | System and method for designing effective business policies via business rules analysis |
| US20100082421A1 (en) * | 2008-09-30 | 2010-04-01 | Yahoo! Inc. | Click through rate prediction system and method |
| US9928552B1 (en) * | 2009-03-13 | 2018-03-27 | New York Life Insurance Company | Methods and systems for insurance investment product decision modeling |
| US20140310053A1 (en) * | 2013-04-10 | 2014-10-16 | Xerox Corporation | Method and systems for providing business process suggestions and recommendations utilizing a business process modeler |
| US20170004513A1 (en) * | 2015-07-01 | 2017-01-05 | Rama Krishna Vadakattu | Subscription churn prediction |
| CN105868298A (en) * | 2016-03-23 | 2016-08-17 | 华南理工大学 | Mobile phone game recommendation method based on binary decision tree |
| WO2017186048A1 (en) * | 2016-04-27 | 2017-11-02 | 第四范式（北京）技术有限公司 | Method and device for presenting prediction model, and method and device for adjusting prediction model |
| WO2017219548A1 (en) * | 2016-06-20 | 2017-12-28 | 乐视控股（北京）有限公司 | Method and device for predicting user attributes |
| CN106817248A (en) * | 2016-12-19 | 2017-06-09 | 西安电子科技大学 | A kind of APT attack detection methods |
| CN107122390A (en) * | 2017-03-04 | 2017-09-01 | 华数传媒网络有限公司 | Recommendation system building method based on groups of users |
| CN107689008A (en) * | 2017-06-09 | 2018-02-13 | 平安科技（深圳）有限公司 | A kind of user insures the method and device of behavior prediction |
| CN107679994A (en) * | 2017-07-26 | 2018-02-09 | 平安科技（深圳）有限公司 | Core protects creation method, device, computer equipment and the storage medium of decision tree |
| CN107679686A (en) * | 2017-08-28 | 2018-02-09 | 阿里巴巴集团控股有限公司 | A kind of business performs method and device |
| CN107730389A (en) * | 2017-09-30 | 2018-02-23 | 平安科技（深圳）有限公司 | Electronic installation, insurance products recommend method and computer-readable recording medium |
| CN107818344A (en) * | 2017-10-31 | 2018-03-20 | 上海壹账通金融科技有限公司 | The method and system that user behavior is classified and predicted |
| CN107944629A (en) * | 2017-11-30 | 2018-04-20 | 北京邮电大学 | A kind of recommendation method and device based on heterogeneous information network representation |
| CN107944990A (en) * | 2017-12-29 | 2018-04-20 | 山大地纬软件股份有限公司 | A kind of integral counter-employee device and method of the precision push based on machine learning |
Non-Patent Citations (2)
| Title |
|---|
| 柯新喜: "基于决策树模型的社会保险客户分类研究", 《福建电脑》 * |
| 柯新喜: "基于决策树模型的社会保险客户分类研究", 《福建电脑》, no. 06, 25 June 2016 (2016-06-25) * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110020728A (en) * | 2018-12-25 | 2019-07-16 | 阿里巴巴集团控股有限公司 | Business model intensified learning method and device |
| CN113965563A (en) * | 2020-12-18 | 2022-01-21 | 深圳平安智汇企业信息管理有限公司 | Model-based business processing method and device and server |
| CN113965563B (en) * | 2020-12-18 | 2023-09-26 | 深圳平安智汇企业信息管理有限公司 | Service processing method and device based on model and server |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108961071B (en) | 2023-07-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104798043B (en) | A kind of data processing method and computer system | |
| CN109885452A (en) | Method for monitoring performance, device and terminal device | |
| CN106650992A (en) | Quantitative investment strategy generating method and apparatus | |
| CN108334954A (en) | Construction method, device, storage medium and the terminal of Logic Regression Models | |
| CN107392645A (en) | Usage mining method, apparatus and its equipment | |
| US20190080248A1 (en) | System and method for facilitating model-based classification of transactions | |
| CN108681751A (en) | Determine the method and terminal device of event influence factor | |
| CN110147389A (en) | Account number treating method and apparatus, storage medium and electronic device | |
| CN109324960A (en) | Automatic test approach and terminal device based on big data analysis | |
| CN108961071A (en) | The method and terminal device of automatic Prediction composite service income | |
| CN111090833A (en) | Data processing method, system and related equipment | |
| CN110796485A (en) | Method and device for improving prediction precision of prediction model | |
| CN110263136B (en) | Method and device for pushing object to user based on reinforcement learning model | |
| CN108304975A (en) | A kind of data prediction system and method | |
| CN115719640B (en) | Chinese medicine primary and secondary symptom recognition system, device, electronic equipment and storage medium thereof | |
| CN107871286A (en) | User is with contacting human world cohesion decision method/system, storage medium and equipment | |
| CN109324963A (en) | The method and terminal device of automatic test profitable result | |
| CN111027669A (en) | Method and device for realizing deep neural network on field programmable gate array | |
| CN116543911A (en) | Disease risk prediction model training method and device | |
| CN114238062B (en) | Board card burning device performance analysis method, device, equipment and readable storage medium | |
| CN115860417A (en) | Agricultural water irrigation distribution method and system and terminal equipment | |
| CN116187675A (en) | Task allocation method, device, equipment and storage medium | |
| KR102406375B1 (en) | An electronic device including evaluation operation of originated technology | |
| CN114627979A (en) | Method and system for determining biomass material characteristic probability distribution information | |
| CN113837368A (en) | Control method and device for evaluating data value of each participant in federal learning |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |