CN110472686A - Object behavior executes probability forecasting method and device - Google Patents
Object behavior executes probability forecasting method and device Download PDFInfo
- Publication number
- CN110472686A CN110472686A CN201910752997.9A CN201910752997A CN110472686A CN 110472686 A CN110472686 A CN 110472686A CN 201910752997 A CN201910752997 A CN 201910752997A CN 110472686 A CN110472686 A CN 110472686A
- Authority
- CN
- China
- Prior art keywords
- attribute information
- behavior
- probability
- similarity
- executes
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- 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"
-
- 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
- G06Q10/10—Office automation; Time management
- G06Q10/105—Human resources
- G06Q10/1053—Employment or hiring
-
- 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
- G06Q40/02—Banking, e.g. interest calculation or account maintenance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Abstract
The present invention provides a kind of object behaviors to execute probability forecasting method and device, this method comprises: obtaining the first attribute information of the first object and the second attribute information of the second object；According to first attribute information and second attribute information, the similarity of the first object and the second object is determined；According to the similarity of the first object and the second object, determine that the first object executes the first probability value of setting behavior；If first probability value is less than predetermined probabilities threshold value, first attribute information is input to behavior probability prediction model, the first object of output executes the second probability value of setting behavior, and the behavior probability prediction model is obtained according to second attribute information.The probability that the present invention can execute setting behavior to object is accurately predicted.
Description
Technical field
The present invention relates to internet areas more particularly to a kind of object behavior to execute probability forecasting method and device.
Background technique
In internet area, if it is possible to predict the subsequent behavior of object, it will data analysis, processing work to object
Generate positive effect.However, not occurring also capableing of effective solution of Accurate Prediction object behavior execution probability in the prior art
Scheme.
Summary of the invention
The embodiment of the present invention proposes that a kind of object behavior executes probability forecasting method, can execute the general of setting behavior to object
Rate accurately predicted, this method comprises:
Obtain the first attribute information of the first object and the second attribute information of the second object；
According to first attribute information and second attribute information, the similar of the first object and the second object is determined
Degree；
According to the similarity of the first object and the second object, determine that the first object executes the first probability value of setting behavior；
If first probability value is less than predetermined probabilities threshold value, first attribute information is input to behavior probability prediction
Model, the first object of output execute the second probability value of setting behavior, and the behavior probability prediction model is according to described second
What attribute information obtained.
The embodiment of the present invention proposes that a kind of object behavior executes probabilistic forecasting device, can execute the general of setting behavior to object
Rate is accurately predicted that the device includes:
Data obtaining module, for obtaining the first attribute information of the first object and the second attribute information of the second object；
Similarity calculation module, for determining first pair according to first attribute information and second attribute information
As the similarity with the second object；
First probabilistic forecasting module determines that the first object executes for the similarity according to the first object and the second object
First probability value of setting behavior；
Second probabilistic forecasting module, if being less than predetermined probabilities threshold value for first probability value, by first attribute
Information input exports the second probability value that the first object executes setting behavior, the behavior probability to behavior probability prediction model
Prediction model is obtained according to second attribute information.
The embodiment of the present invention also proposed a kind of computer equipment, including memory, processor and storage are on a memory
And the computer program that can be run on a processor, the processor realize above-mentioned object behavior when executing the computer program
Execute probabilistic method.
The embodiment of the present invention also proposed a kind of computer readable storage medium, the computer-readable recording medium storage
Have and executes the computer program that above-mentioned object behavior executes probabilistic method.
In embodiments of the present invention, the first attribute information of the first object and the second attribute information of the second object are obtained；
According to first attribute information and second attribute information, the similarity of the first object and the second object is determined；According to
The similarity of an object and the second object determines that the first object executes the first probability value of setting behavior；If first probability
Value is less than predetermined probabilities threshold value, first attribute information is input to behavior probability prediction model, the first object of output executes
Second probability value of setting behavior, the behavior probability prediction model are obtained according to second attribute information.Above-mentioned
In the process, first according to the second attribute information of the first attribute information of the first object and the second object, it is determined that the first object
With the similarity of the second object, so that the first probability that the first object executes pending setting behavior is obtained, in the first probability
When less than setting probability threshold value, only the first attribute information of the first object need to be input to behavior probability prediction model, can obtained
The second probability that the first object executes pending setting behavior is obtained, above-mentioned prediction process considers the second attribute letter of the second object
Breath, predictablity rate are high.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
It obtains other drawings based on these drawings.In the accompanying drawings:
Fig. 1 is the flow chart that object behavior executes probability forecasting method in the embodiment of the present invention；
Fig. 2 is the detail flowchart that the object behavior that the embodiment of the present invention proposes executes probability forecasting method；
Fig. 3 is the schematic diagram that object behavior executes probabilistic forecasting device in the embodiment of the present invention.
Specific embodiment
Understand in order to make the object, technical scheme and advantages of the embodiment of the invention clearer, with reference to the accompanying drawing to this hair
Bright embodiment is described in further details.Here, the illustrative embodiments of the present invention and their descriptions are used to explain the present invention, but simultaneously
It is not as a limitation of the invention.
In internet area, if it is possible to predict the subsequent behavior of object, it will data analysis, processing work to object
Generate positive effect.For example, in bank's recruitment, if it is possible to predict that applicant stays in the probability of bank's work, then may be used
The a large amount of recruitment cost of bank and time are saved, the efficiency of recruitment is improved and recruits the accuracy rate of the suitable talent.
Fig. 1 is the flow chart that object behavior executes probability forecasting method in the embodiment of the present invention, as shown in Figure 1, this method
Include:
Step 101, the first attribute information of the first object and the second attribute information of the second object are obtained, wherein first
Object is the object of pending setting behavior, and the second object is the object for having executed setting behavior；
Step 102, according to first attribute information and second attribute information, the first object and the second object are determined
Similarity；
Step 103, according to the similarity of the first object and the second object, determine that the first object executes the first of setting behavior
Probability value；
Step 104, if first probability value is less than predetermined probabilities threshold value, first attribute information is input to behavior
Probabilistic Prediction Model, the first object of output execute the second probability value of setting behavior, and the behavior probability prediction model is basis
What second attribute information obtained.
In embodiments of the present invention, the first attribute information of the first object and the second attribute information of the second object are obtained；
According to first attribute information and second attribute information, the similarity of the first object and the second object is determined；According to
The similarity of an object and the second object determines that the first object executes the first probability value of setting behavior；If first probability
Value is less than predetermined probabilities threshold value, first attribute information is input to behavior probability prediction model, the first object of output executes
Second probability value of setting behavior, the behavior probability prediction model are obtained according to second attribute information.Above-mentioned
In the process, first according to the second attribute information of the first attribute information of the first object and the second object, it is determined that the first object
With the similarity of the second object, so that the first probability that the first object executes pending setting behavior is obtained, in the first probability
When less than setting probability threshold value, only the first attribute information of the first object need to be input to behavior probability prediction model, can obtained
The second probability that the first object executes pending setting behavior is obtained, above-mentioned prediction process considers the second attribute letter of the second object
Breath, predictablity rate are high.
In a step 101, the first attribute information of the first object can there are many forms, by taking bank recruits as an example, first
Object is the object of pending setting behavior, can be applicant, and setting behavior stays in bank's work, and the second object is to have held
The object of row setting behavior, can be bank employee, then, the first attribute information of applicant includes gender, political affiliation, nationality
It passes through, the age, working experience, address, marital status, education (education of undergraduate course, graduate education, doctor education), profession, whether obtain
Obtain English Band certificate, whether obtain ENGLISH BAND SIX CERTIFICATE, whether obtain finance and economic certificate etc.；The attribute information of bank employee can
To include gender, political affiliation, native place, age, working experience, address, marital status, education (education of undergraduate course, postgraduate's religion
Educate, doctor education), profession, whether obtain English Band certificate, whether obtain ENGLISH BAND SIX CERTIFICATE, whether obtain finance and economic card
Book, the total working time limit, bank's length of service, nearly 3 years job performances, current level, evaluation informations etc..
In a step 102, the phase of above-mentioned first attribute information and the second attributive analysis the first object and the second attribute is utilized
Like degree, if two object similarities are high, since the second object is the object for having executed setting behavior, then the first object is held
The probability of row setting behavior is with regard to bigger, by taking bank recruits as an example, if the similarity-rough set of applicant and bank employee are big,
Applicant stays in the probability of bank's work with regard to big, subsequent that emphasis is arranged to interview or preferential interview for such applicant, because
This, enters step 103, according to the similarity of the first object and the second object, determines that the first object executes the first of setting behavior
Probability value can carry out first time prediction to the first object in this way, improve prediction accuracy.
At step 104, if being less than predetermined probabilities threshold value according to the first probability value that similarity determines, for example, application
The similarity of person and bank employee are smaller, and the probability that the applicant thereby determined that stays in bank can be less than predetermined probabilities threshold value, but
Applicant can not say that the applicant is not suitable for staying in bank when smaller with the similarity of bank employee, therefore, it is also desirable into
First attribute information is input to behavior probability prediction model by one step, and the second of output the first object execution setting behavior is general
Rate value, and behavior probability prediction model is obtained according to second attribute information, therefore, entire prediction process considers
The attribute information of the object of setting behavior is executed, prediction accuracy is high.
In one embodiment, according to first attribute information and second attribute information, determine the first object and
Before the similarity of second object, further includes:
Second attribute information of the first attribute information and the second object to the first object is digitally converted processing, obtains
To corresponding first digitized result of the first attribute information and corresponding second digitized result of the second attribute information；
According to first attribute information and second attribute information, the similar of the first object and the second object is determined
Degree, comprising:
According to first digitized result and second digitized result, the phase of the first object and the second object is determined
Like degree.
In the above-described embodiments, the second attribute information of the first attribute information of the first object and the second object is counted
Word conversion process, the result of acquisition are conducive to calculate similarity using means such as mathematics publicities, can be to the first object
Second attribute information of the first attribute information and the second object is converted to symbol measurement or Boolean, and part first is shown below
The digitized processing process of attribute information:
Gender: male value 1, female's value 0；
Political affiliation: party member's value 2, non-Party member's value 3；
Native place: a line city value 4, tier 2 cities value 5, three line city values 6；
Age: less than 32 values 7 is greater than 32 values 8；
Working experience: there are bank's working experience value 9, no bank's working experience value 10；
Address: working road duration value 11 within 1 hour, working road duration are greater than 1 hour value 12；
Marital status: unmarried value 13, married value 14；
Education: education of undergraduate course takes 15, and graduate education takes 16, and doctor education takes 17；
Profession: financial class profession value 18, non-financial class profession value 19；
Whether obtain English Band certificate: obtaining English Band certificate value 20, do not obtain English Band certificate value 21；
Whether obtain ENGLISH BAND SIX CERTIFICATE: obtaining ENGLISH BAND SIX CERTIFICATE value 22, do not obtain ENGLISH BAND SIX CERTIFICATE value 23；
Whether obtain finance and economic certificate: obtaining finance and economic certificate value 24, does not obtain and obtain finance and economic certificate value 25.
In one embodiment, it is determined using following formula according to first attribute information and second attribute information
The similarity of first object and the second object:
Wherein, Jaccard (x, y) is the similarity of the first object and the second object；
X is the first attribute information of the first object；
Y is the second attribute information of the second object.
In the above-described embodiments, Jaccard coefficient is mainly used between the object for calculating symbol measurement or boolean's value metric
Similarity, because the attribute information of object is all that can not measure difference occurrence by symbol measurement or Boolean mark
Size, can only obtain " whether identical " this as a result, so Jaccard coefficient has the characteristic that jointly between being only concerned object
No consistent this problem, if the Jaccard similarity factor of comparison other X and object Y, comparison other X is identical with object Y
Number.Such as applicant the first attribute information and bank employee the second attribute information in all include gender, political affiliation,
Native place, the age, working experience, address, marital status, education (education of undergraduate course, graduate education, doctor education), profession, whether
English Band certificate is obtained, ENGLISH BAND SIX CERTIFICATE whether is obtained, whether obtains finance and economic certificate, wherein the first of applicant belongs to
Property information in X gender, political affiliation, native place, the age, working experience, address, marital status, education, profession, if obtain English
Language level Four certificate, if obtain ENGLISH BAND SIX CERTIFICATE, if obtain finance and economic certificate }=male, and party member, Beijing, 24, no bank
Working experience, Beijing Haidian, unmarried, postgraduate, finance obtain English Band certificate, obtain ENGLISH BAND SIX CERTIFICATE, do not obtain wealth
Through class certificate }, above-mentioned X is digitally converted and obtains X={ 1,2,4,7,10,11,13,16,18,20,22,25 }, it is a certain
X { gender, political affiliation, native place, age, working experience, address, marital status, religion in the second attribute information of bank employee
Educate, profession, if obtain English Band certificate, if obtain ENGLISH BAND SIX CERTIFICATE, if obtain finance and economic certificate={ male, party
Member, Beijing, 35, there are bank's working experience, Beijing Haidian, married, doctor, finance, acquisition English Band certificate, acquisition English six
Grade certificate, obtain finance and economic certificate, by above-mentioned Y be digitally converted obtain Y=1,2,4,8,9,11,14,16,18,20,
22,24 }, wherein can then be obtained according to formula (1) in X and Y:
When there are multiple second objects, the similarity of the first object He each the second object is calculated separately, to obtain
Multiple similarities are obtained, then takes the average value of multiple similarities as the similarity of the first object and the second object, can mention in this way
The accuracy of high first object and the second object similarity calculation.
When it is implemented, determining that the first object executes setting behavior according to the similarity of the first object and the second object
One of embodiment is given below in there are many ways to first probability value.
In one embodiment, according to the similarity of the first object and the second object, determine that the first object executes setting behavior
The first probability value, comprising:
Determine setting similarity threshold；
If the similarity of the first object and the second object is greater than setting similarity threshold, it is pending to determine that the first object executes
The first probability value of behavior is set as the first probability setting value.
In the above-described embodiments, setting similarity threshold can be preconfigured, such as can be 0.5, the first probability
Given threshold can be preconfigured to 0.6, certainly, i.e., when the similarity of the first object and the second object is greater than 0.5, determine first
Object executes pending the first probability value for setting behavior as the first probability setting value, and as 0.6.
When it is implemented, predetermined probabilities threshold value can be 0.6, the predetermined probabilities threshold value and the first probability setting value be not straight
Relationship is connect, is less than predetermined probabilities threshold value in the first probability value, first attribute information is input to behavior probability prediction model,
The second probability value that the first object executes setting behavior is exported, the behavior probability prediction model is believed according to second attribute
What breath obtained.
In one embodiment, behavior probability prediction model includes that object behavior disaggregated model and probability obtain model；
Behavior probability prediction model obtains with the following method:
According to the second attribute information of the second object, training obtains object behavior disaggregated model, the object behavior classification
Model is used to obtain the behavior classification results of the first object；
It constructs probability and obtains model, the probability obtains model for the behavior classification results according to the first object, obtains
First object executes the second probability of pending setting behavior.
According to the second attribute information of the second object, training obtains object behavior disaggregated model, comprising:
Extract the feature vector of the second attribute information of the second object；
Utilize described eigenvector training object behavior disaggregated model.
In the above-described embodiments, object behavior disaggregated model is a kind of machine learning model, in one embodiment, object row
It can be that gradient promotes decision tree (Gradient Boosting Decision Tree, GBDT) model, GBDT for disaggregated model
Model is a kind of decision Tree algorithms model of iteration, which is made of more decision trees, and the conclusion of all trees adds up
To do final result.In the iteration of GBDT model, it is assumed that the strong learner that previous round iteration obtains is ft-1(x), loss function
It is L (y, ft-1(x)), the target of epicycle iteration is to find a weak learner ht(x), loss function L (y, the f of epicycle are allowedt-1
(x)+ht(x)) minimum.That is, epicycle iteration finds decision tree, the loss of sample to be allowed to become smaller as far as possible.Above-mentioned
In embodiment, it may include that the first order (answer by low significance level so that bank recruits as an example that behavior classification results, which may include a variety of,
The person of engaging), the second level (middle significance level applicant) and the third level (high significance level applicant), then probability obtain model be used for
According to the behavior classification results of the first object, the second probability that the first object executes pending setting behavior is obtained, for example, answering
When the person of engaging is first order applicant, the probability that applicant stays in bank's work is 30%, when applicant is second level applicant,
The probability that applicant stays in bank's work is 60%, and when applicant is third level applicant, applicant stays in bank's work
Probability is 90%.
Based on the above embodiment, the present invention proposes following one embodiment to illustrate that object behavior executes probability forecasting method
Detailed process, Fig. 2 be the embodiment of the present invention propose object behavior execute probability forecasting method detail flowchart, such as Fig. 2
Shown, in one embodiment, the detailed process that object behavior executes probability forecasting method includes:
Step 201, the first attribute information of the first object and the second attribute information of the second object are obtained, wherein first
Object is the object of pending setting behavior, and the second object is the object for having executed setting behavior；
Step 202, digitlization is carried out to the second attribute information of the first attribute information of the first object and the second object to turn
Processing is changed, corresponding first digitized result of the first attribute information and corresponding second digitlization of the second attribute information are obtained
As a result；
Step 203, according to first digitized result and second digitized result, the first object and second are determined
The similarity of object；
Step 204, setting similarity threshold is determined；
Step 205, if the similarity of the first object and the second object is greater than setting similarity threshold, determine that the first object is held
Pending the first probability value for setting behavior of row is the first probability setting value；
Step 206, if first probability value is less than predetermined probabilities threshold value, first attribute information is input to behavior
Probabilistic Prediction Model, the first object of output execute the second probability value of setting behavior, and the behavior probability prediction model is basis
What second attribute information obtained.
It is, of course, understood that the detailed process that above-mentioned object behavior executes probability forecasting method can also have other
Change case, associated change example should all fall into protection scope of the present invention.
The embodiment of the present invention propose method in, obtain the first object the first attribute information and the second object second
Attribute information；According to first attribute information and second attribute information, the similar of the first object and the second object is determined
Degree；According to the similarity of the first object and the second object, determine that the first object executes the first probability value of setting behavior；If described
First probability value is less than predetermined probabilities threshold value, and first attribute information is input to behavior probability prediction model, output first
Object executes the second probability value of setting behavior, and the behavior probability prediction model is obtained according to second attribute information
's.In above process, first according to the second attribute information of the first attribute information of the first object and the second object, it is determined that
The similarity of first object and the second object, so that the first probability that the first object executes pending setting behavior is obtained, In
When first probability is less than setting probability threshold value, the first attribute information of the first object need to be only input to behavior probability prediction mould
Type, can be obtained the second probability that the first object executes pending setting behavior, and above-mentioned prediction process considers the second object
Second attribute information, predictablity rate are high.
In addition, the similarity of the first object He each the second object is calculated separately when there are multiple second objects, from
And multiple similarities are obtained, then take the average value of multiple similarities as the similarity of the first object and the second object, in this way
The accuracy of the first object and the second object similarity calculation can be improved.
Based on same inventive concept, the embodiment of the invention also provides a kind of object behaviors to execute probabilistic forecasting device,
As described in the following examples.Since the principle that these are solved the problems, such as is similar to object behavior execution probability forecasting method,
The implementation of device may refer to the implementation of method, repeats place and is not repeating.
Fig. 3 is the schematic diagram that object behavior executes probabilistic forecasting device in the embodiment of the present invention, as shown in figure 3, the device
Include:
Data obtaining module 301, for obtaining the first attribute information of the first object and the second attribute letter of the second object
Breath；
Similarity calculation module 302, for determining first according to first attribute information and second attribute information
The similarity of object and the second object；
First probabilistic forecasting module 303 determines that the first object is held for the similarity according to the first object and the second object
First probability value of row setting behavior；
Second probabilistic forecasting module 304 belongs to if being less than predetermined probabilities threshold value for first probability value by described first
Property information input to behavior probability prediction model, export the second probability value that the first object executes setting behavior, the behavior is general
Rate prediction model is obtained according to second attribute information.
In one embodiment, object behavior executes probabilistic forecasting device, further includes digitlization conversion processing module 305, uses
In:
Second attribute information of the first attribute information and the second object to the first object is digitally converted processing, obtains
To corresponding first digitized result of the first attribute information and corresponding second digitized result of the second attribute information；
Similarity calculation module 302 is specifically used for:
According to first digitized result and second digitized result, the phase of the first object and the second object is determined
Like degree.
In one embodiment, the first probabilistic forecasting module 303 is specifically used for:
Determine setting similarity threshold；
If the similarity of the first object and the second object is greater than setting similarity threshold, it is pending to determine that the first object executes
The first probability value of behavior is set as the first probability setting value.
In one embodiment, behavior probability prediction model includes that object behavior disaggregated model and probability obtain model；
Behavior probability prediction model obtains with the following method:
According to the second attribute information of the second object, training obtains object behavior disaggregated model, the object behavior classification
Model is used to obtain the behavior classification results of the first object；
It constructs probability and obtains model, the probability obtains model for the behavior classification results according to the first object, obtains
First object executes the second probability of pending setting behavior.
In one embodiment, behavior probability prediction model specifically obtains with the following method:
Extract the feature vector of the second attribute information of the second object；
Utilize described eigenvector training object behavior disaggregated model.
The embodiment of the present invention propose method in, obtain the first object the first attribute information and the second object second
Attribute information；According to first attribute information and second attribute information, the similar of the first object and the second object is determined
Degree；According to the similarity of the first object and the second object, determine that the first object executes the first probability value of setting behavior；If described
First probability value is less than predetermined probabilities threshold value, and first attribute information is input to behavior probability prediction model, output first
Object executes the second probability value of setting behavior, and the behavior probability prediction model is obtained according to second attribute information
's.In above process, first according to the second attribute information of the first attribute information of the first object and the second object, it is determined that
The similarity of first object and the second object, so that the first probability that the first object executes pending setting behavior is obtained, In
When first probability is less than setting probability threshold value, the first attribute information of the first object need to be only input to behavior probability prediction mould
Type, can be obtained the second probability that the first object executes pending setting behavior, and above-mentioned prediction process considers the second object
Second attribute information, predictablity rate are high.
In addition, the similarity of the first object He each the second object is calculated separately when there are multiple second objects, from
And multiple similarities are obtained, then take the average value of multiple similarities as the similarity of the first object and the second object, in this way
The accuracy of the first object and the second object similarity calculation can be improved.
It should be understood by those skilled in the art that, the embodiment of the present invention can provide as method, system or computer program
Product.Therefore, complete hardware embodiment, complete software embodiment or reality combining software and hardware aspects can be used in the present invention
Apply the form of example.Moreover, it wherein includes the computer of computer usable program code that the present invention, which can be used in one or more,
The computer program implemented in usable storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.) produces
The form of product.
The present invention be referring to according to the method for the embodiment of the present invention, the process of equipment (system) and computer program product
Figure and/or block diagram describe.It should be understood that every one stream in flowchart and/or the block diagram can be realized by computer program instructions
The combination of process and/or box in journey and/or box and flowchart and/or the block diagram.It can provide these computer programs
Instruct the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce
A raw machine, so that being generated by the instruction that computer or the processor of other programmable data processing devices execute for real
The device for the function of being specified in present one or more flows of the flowchart and/or one or more blocks of the block diagram.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
Particular embodiments described above has carried out further in detail the purpose of the present invention, technical scheme and beneficial effects
Describe in detail it is bright, it should be understood that the above is only a specific embodiment of the present invention, the guarantor being not intended to limit the present invention
Range is protected, all within the spirits and principles of the present invention, any modification, equivalent substitution, improvement and etc. done should be included in this
Within the protection scope of invention.
Claims (10)
1. a kind of object behavior executes probability forecasting method characterized by comprising
Obtain the first attribute information of the first object and the second attribute information of the second object, wherein the first object is pending
The object of setting behavior, the second object are the object for having executed setting behavior；
According to first attribute information and second attribute information, the similarity of the first object and the second object is determined；
According to the similarity of the first object and the second object, determine that the first object executes the first probability value of setting behavior；
If first probability value is less than predetermined probabilities threshold value, first attribute information is input to behavior probability prediction mould
Type, the first object of output execute the second probability value of setting behavior, and the behavior probability prediction model is belonged to according to described second
Property information acquisition.
2. object behavior as described in claim 1 executes probability forecasting method, which is characterized in that according to first attribute
Information and second attribute information, before the similarity for determining the first object and the second object, further includes:
Second attribute information of the first attribute information and the second object to the first object is digitally converted processing, obtains
Corresponding first digitized result of one attribute information and corresponding second digitized result of the second attribute information；
According to first attribute information and second attribute information, the similarity of the first object and the second object is determined, wrap
It includes:
According to first digitized result and second digitized result, the similar of the first object and the second object is determined
Degree.
3. object behavior as described in claim 1 executes probability forecasting method, which is characterized in that according to the first object and second
The similarity of object determines that the first object executes the first probability value of setting behavior, comprising:
Determine setting similarity threshold；
If the similarity of the first object and the second object is greater than setting similarity threshold, determine that the first object executes pending setting
First probability value of behavior is the first probability setting value.
4. object behavior as described in claim 1 executes probability forecasting method, which is characterized in that behavior probability prediction model packet
It includes object behavior disaggregated model and probability obtains model；
Behavior probability prediction model obtains with the following method:
According to the second attribute information of the second object, training obtains object behavior disaggregated model, the object behavior disaggregated model
For obtaining the behavior classification results of the first object；
It constructs probability and obtains model, the probability obtains model for the behavior classification results according to the first object, obtains first
Object executes the second probability of pending setting behavior.
5. object behavior as claimed in claim 4 executes probability forecasting method, which is characterized in that according to the second of the second object
Attribute information, training obtain object behavior disaggregated model, comprising:
Extract the feature vector of the second attribute information of the second object；
Utilize described eigenvector training object behavior disaggregated model.
6. object behavior as claimed in claim 2 executes probability forecasting method, which is characterized in that following formula is used, according to
First attribute information and second attribute information, determine the similarity of the first object and the second object:
Wherein, Jaccard (x, y) is the similarity of the first object and the second object；
X is the first attribute information of the first object；
Y is the second attribute information of the second object.
7. a kind of object behavior executes probabilistic forecasting device characterized by comprising
Data obtaining module, for obtaining the first attribute information of the first object and the second attribute information of the second object；
Similarity calculation module, for according to first attribute information and second attribute information, determine the first object and
The similarity of second object；
First probabilistic forecasting module determines that the first object executes setting for the similarity according to the first object and the second object
First probability value of behavior；
Second probabilistic forecasting module, if being less than predetermined probabilities threshold value for first probability value, by first attribute information
It is input to behavior probability prediction model, the first object of output executes the second probability value of setting behavior, the behavior probability prediction
Model is obtained according to second attribute information.
8. object behavior as claimed in claim 7 executes probabilistic forecasting device, which is characterized in that further include at digitlization conversion
Module is managed, is used for:
Second attribute information of the first attribute information and the second object to the first object is digitally converted processing, obtains
Corresponding first digitized result of one attribute information and corresponding second digitized result of the second attribute information；
Similarity calculation module is specifically used for:
According to first digitized result and second digitized result, the similar of the first object and the second object is determined
Degree.
9. a kind of computer equipment including memory, processor and stores the meter that can be run on a memory and on a processor
Calculation machine program, which is characterized in that the processor is realized described in any one of claim 1 to 6 when executing the computer program
Method.
10. a kind of computer readable storage medium, which is characterized in that the computer-readable recording medium storage has perform claim
It is required that the computer program of any one of 1 to 6 the method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910752997.9A CN110472686A (en) | 2019-08-15 | 2019-08-15 | Object behavior executes probability forecasting method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910752997.9A CN110472686A (en) | 2019-08-15 | 2019-08-15 | Object behavior executes probability forecasting method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110472686A true CN110472686A (en) | 2019-11-19 |
Family
ID=68510842
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910752997.9A Pending CN110472686A (en) | 2019-08-15 | 2019-08-15 | Object behavior executes probability forecasting method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110472686A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105095588A (en) * | 2015-08-05 | 2015-11-25 | 中国联合网络通信集团有限公司 | Mobile internet user complaint prediction method and device |
| CN107203866A (en) * | 2017-06-26 | 2017-09-26 | 北京京东尚科信息技术有限公司 | The processing method and device of order |
| CN108932525A (en) * | 2018-06-07 | 2018-12-04 | 阿里巴巴集团控股有限公司 | A kind of behavior prediction method and device |
| CN109345376A (en) * | 2018-09-27 | 2019-02-15 | 北京芯盾时代科技有限公司 | A kind of e-bank is counter to cheat method and system |
| CN109858970A (en) * | 2019-02-02 | 2019-06-07 | 中国银行股份有限公司 | A kind of user's behavior prediction method, apparatus and storage medium |
| CN109902849A (en) * | 2018-06-20 | 2019-06-18 | 华为技术有限公司 | User's behavior prediction method and device, behavior prediction model training method and device |
-
2019
- 2019-08-15 CN CN201910752997.9A patent/CN110472686A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105095588A (en) * | 2015-08-05 | 2015-11-25 | 中国联合网络通信集团有限公司 | Mobile internet user complaint prediction method and device |
| CN107203866A (en) * | 2017-06-26 | 2017-09-26 | 北京京东尚科信息技术有限公司 | The processing method and device of order |
| CN108932525A (en) * | 2018-06-07 | 2018-12-04 | 阿里巴巴集团控股有限公司 | A kind of behavior prediction method and device |
| CN109902849A (en) * | 2018-06-20 | 2019-06-18 | 华为技术有限公司 | User's behavior prediction method and device, behavior prediction model training method and device |
| CN109345376A (en) * | 2018-09-27 | 2019-02-15 | 北京芯盾时代科技有限公司 | A kind of e-bank is counter to cheat method and system |
| CN109858970A (en) * | 2019-02-02 | 2019-06-07 | 中国银行股份有限公司 | A kind of user's behavior prediction method, apparatus and storage medium |
Non-Patent Citations (2)
| Title |
|---|
| 张雪英: "《情报检索语言的兼容转换》", 31 December 2009, 东南大学出版社 * |
| 高永彬等: "《Hadoop大数据分析》", 31 July 2019, 中国铁道出版社 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113822494B (en) | Risk prediction method, device, equipment and storage medium | |
| US20230325724A1 (en) | Updating attribute data structures to indicate trends in attribute data provided to automated modelling systems | |
| CN111783875A (en) | Abnormal user detection method, device, equipment and medium based on cluster analysis | |
| JP6414363B2 (en) | Prediction system, method and program | |
| CN111597348B (en) | User image drawing method, device, computer equipment and storage medium | |
| US11210673B2 (en) | Transaction feature generation | |
| CN115082920A (en) | Deep learning model training method, image processing method and device | |
| CN110782349A (en) | Model training method and system | |
| CN112041880A (en) | Deep learning method for assessing credit risk | |
| CN113706291A (en) | Fraud risk prediction method, device, equipment and storage medium | |
| CN113886697A (en) | Clustering algorithm based activity recommendation method, device, equipment and storage medium | |
| CN110213239B (en) | Suspicious transaction message generation method and device and server | |
| CN111353728A (en) | Risk analysis method and system | |
| WO2019218517A1 (en) | Server, method for processing text data and storage medium | |
| CN110544166A (en) | Sample generation method, device and storage medium | |
| CN112463964B (en) | Text classification and model training method, device, equipment and storage medium | |
| CN110472686A (en) | Object behavior executes probability forecasting method and device | |
| CN113706258A (en) | Product recommendation method, device, equipment and storage medium based on combined model | |
| CN113591881A (en) | Intention recognition method and device based on model fusion, electronic equipment and medium | |
| CN114418776A (en) | Data processing method, device, terminal equipment and medium | |
| CA3097731A1 (en) | System and method for deep learning recommender | |
| CN111898708A (en) | Transfer learning method and electronic equipment | |
| KR20210035622A (en) | Time series data similarity calculation system and method | |
| CN111080433A (en) | Credit risk assessment method and device | |
| CN110223102A (en) | Lead referral method, apparatus, electronic equipment and storage medium |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191119 |
|
| RJ01 | Rejection of invention patent application after publication |