CN108280542A - A kind of optimization method, medium and the equipment of user's portrait model - Google Patents
A kind of optimization method, medium and the equipment of user's portrait model Download PDFInfo
- Publication number
- CN108280542A CN108280542A CN201810035915.4A CN201810035915A CN108280542A CN 108280542 A CN108280542 A CN 108280542A CN 201810035915 A CN201810035915 A CN 201810035915A CN 108280542 A CN108280542 A CN 108280542A
- Authority
- CN
- China
- Prior art keywords
- data
- prediction
- prediction model
- training
- model
- 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
- 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
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
- G06Q30/0203—Market surveys; Market polls
-
- 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
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0255—Targeted advertisements based on user history
-
- 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
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0269—Targeted advertisements based on user profile or attribute
- G06Q30/0271—Personalized advertisement
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Physics & Mathematics (AREA)
- Marketing (AREA)
- Game Theory and Decision Science (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Data Mining & Analysis (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention provides optimization method, medium and the equipment of a kind of user portrait model.The method, including：Obtain user behavior data；The first prediction result is obtained based on the first prediction model pre-established according to the behavioral data；According to first prediction result and the behavioral data, training the second prediction model of optimization.By obtaining user behavior data, according to the first prediction model, obtain the first prediction result, and according to the first prediction result and behavioral data, training the second prediction model of optimization, by increasing the input sample data type of training Optimized model to the second prediction model, the accuracy of the second prediction model prediction can be improved；Meanwhile the second prediction model of optimization is trained by using the prediction result of the first prediction model, when the first prediction model changes, it can realize the automatic training optimization to the second prediction model, and then the time can be saved, reduce cost.
Description
Technical field
The present invention relates to big data machine learning fields, and in particular to a kind of optimization method, the medium of user's portrait model
And equipment.
Background technology
Under the background of big data, it would be desirable to according to the labels such as the behavior of user, gender, age carry out it is purposeful, point
The advertisement pushing of the Products Show of classification is segmented, the purpose of precision marketing with realizing that user draws a portrait.Not with Internet technology
Disconnected upgrading, when we establish user using user data draws a portrait, user can receive various information all the time, often
The selection made in a flash is all different, it is difficult to be analyzed user with a fixed model and behavior prediction.This
Sample just loses real-time, causes message lag, the phenomenon of prediction result inaccuracy, causes marketing effectiveness poor, advertisement pushes
Conversion ratio is low.If being required for artificial modification model every time, manually if update, no matter from time cost or human cost
All it is huge for upper, and effect is also bad.
Invention content
For the defects in the prior art, the present invention, which provides a kind of user and draws a portrait, the optimization method of model, medium and sets
It is standby, the accuracy of prediction model can be improved, the model optimization time can be saved, reduces cost.
In a first aspect, a kind of optimization method for model of drawing a portrait the present invention provides user, including：
Obtain user behavior data；
The first prediction result is obtained based on the first prediction model pre-established according to the behavioral data；
According to first prediction result and the behavioral data, training the second prediction model of optimization.
Optionally, the first prediction is being obtained based on the first prediction model pre-established according to first behavioral data
As a result before the step of, further include：
OF17-P19424
Obtain sample data；
Classify to the sample data；
Data cleansing is carried out to the sorted sample data, obtains the feature samples data of the sample data；
According to the feature samples data, the first prediction model of training.
Optionally, described to classify to the sample data, including：
According to behavior pattern, classify to the sample data.
Optionally, according to the feature samples data, the step of the first prediction model of training before, further include：
The feature samples data are combined, new feature sample data is obtained；
It is described to train the first prediction model according to the feature samples data, including：
According to the new feature sample data, the first prediction model of training.
Optionally, it after the step of being combined to the feature samples data, obtaining new feature sample data, also wraps
It includes：
The new feature sample data is normalized；
It is described to train the first prediction model according to the feature samples data, including：
According to the new feature sample data after normalized, the first prediction model of training.
Optionally, after the step of training the first prediction model, further include：
Obtain test data；
According to the test data, the accuracy score value of first prediction model is calculated；
Judge whether the accuracy score value is more than default accuracy threshold value；
If more than then executing described according to the behavioral data, based on the first prediction model pre-established, obtain first
The step of prediction result；
If being not more than, re-execute it is described according to the feature samples data, the step of the first prediction model of training.
Optionally, according to first prediction result and the behavioral data, the step of training the second prediction model of optimization
Before rapid, further include：
Based on first prediction model, the confidence level of first prediction result is obtained；
Judge whether the confidence level is less than corresponding first threshold；If being less than, re-execute described according to the spy
The step of levying sample data, training the first prediction model；
If being not less than, first prediction result is exported；And judge whether the confidence level is more than corresponding second threshold
Value；
If more than then judging that first prediction result can be used as training characteristics；If being not more than, judge that described first is pre-
It surveys result and not can be used as training characteristics；
It is described that according to first prediction result and the behavioral data, training optimizes the second prediction model, including：
It is more than first prediction result of the second threshold and the behavioral data according to the confidence level, training is excellent
Change the second prediction model.
Second aspect, the present invention provide a kind of computer readable storage medium, are stored thereon with computer program, the program
A kind of optimization method of above-mentioned user's portrait model is realized when being executed by processor.
The third aspect, the present invention provide a kind of optimization equipment of user's portrait model, including：It memory, processor and deposits
The computer program that can be run on a memory and on a processor is stored up, the processor realizes above-mentioned one when executing described program
The optimization method of kind user's portrait model.
The present invention provides a kind of optimization methods of user portrait model, including：Obtain user behavior data；According to described
Behavioral data obtains the first prediction result based on the first prediction model pre-established；According to first prediction result and institute
State behavioral data, training the second prediction model of optimization.By obtaining user behavior data, according to the first prediction model, the is obtained
One prediction result, and according to the first prediction result and behavioral data, training the second prediction model of optimization, by giving the second prediction mould
Type increases the input sample data type of training Optimized model, can improve the accuracy of the second prediction model prediction；Meanwhile it is logical
Cross prediction result training the second prediction model of optimization using the first prediction model, when the first prediction model changes, Neng Goushi
Now the automatic training of the second prediction model is optimized, and then the time can be saved, reduces cost.
A kind of computer readable storage medium provided by the invention and a kind of user draw a portrait the optimization equipment of model, and above-mentioned
The optimization method of user's portrait model is for identical inventive concept, advantageous effect having the same.
Description of the drawings
It, below will be to specific in order to illustrate more clearly of the specific embodiment of the invention or technical solution in the prior art
Embodiment or attached drawing needed to be used in the description of the prior art are briefly described.In all the appended drawings, similar element
Or part is generally identified by similar reference numeral.In attached drawing, each element or part might not be drawn according to actual ratio.
Fig. 1 is a kind of flow chart of the optimization method of user's portrait model provided by the invention；
Fig. 2 is a kind of structural schematic diagram of the optimization equipment of user's portrait model provided by the invention.
Specific implementation mode
The embodiment of technical solution of the present invention is described in detail below in conjunction with attached drawing.Following embodiment is only used for
Clearly illustrate technical scheme of the present invention, therefore be intended only as example, and the protection of the present invention cannot be limited with this
Range.
It should be noted that unless otherwise indicated, technical term or scientific terminology used in this application should be this hair
The ordinary meaning that bright one of ordinary skill in the art are understood.
The present invention provides optimization method, medium and the equipment of a kind of user portrait model.Below in conjunction with the accompanying drawings to this
The embodiment of invention illustrates.
First embodiment：
The signal of the optimization method of model referring to FIG. 1, a kind of user that Fig. 1, which is the specific embodiment of the invention, to be provided draws a portrait
Figure, a kind of optimization method of user's portrait model provided in this embodiment, including：
Step S101：Obtain user behavior data.
Step S102：The first prediction knot is obtained based on the first prediction model pre-established according to the behavioral data
Fruit.
Step S103：According to first prediction result and the behavioral data, training the second prediction model of optimization.
By obtaining user behavior data, according to the first prediction model, the first prediction result is obtained, and according to the first prediction
As a result and behavioral data, training the second prediction model of optimization train the input of Optimized model by increasing to the second prediction model
Sample data type can improve the accuracy of the second prediction model prediction；Meanwhile by using the prediction of the first prediction model
As a result the second prediction model of training optimization can realize the automatic instruction to the second prediction model when the first prediction model changes
Practice optimization, and then the time can be saved, reduces cost.
In the present invention, the first prediction model can be multiple prediction models, can also be a prediction model；Second is pre-
It can be multiple prediction models to survey model, can also be a prediction model.Here the first prediction model and second are not intended to limit
The number of prediction model, all within the scope of the present invention.
For example, can be excellent as the training of multiple second prediction models using the first prediction result of multiple first prediction models
Change data.
In a specific embodiment provided by the invention, according to first behavioral data, based on what is pre-established
Before the step of first prediction model, the first prediction result of acquisition, further include：Obtain sample data；To the sample data into
Row classification；Data cleansing is carried out to the sorted sample data, obtains the feature samples data of the sample data；According to
The feature samples data, the first prediction model of training.
Before obtaining the first prediction result using the first prediction model, should also include：The first prediction model of training.
The process of the first prediction model of training is as follows：
The first step obtains sample data, wherein sample data includes input sample data and output sample data.Wherein,
Sample data can be user behavior data.For example, being mounted with any APP, opening which APP, which place gone to, is lived
The room rate etc. of ground cell.
Second step classifies to sample data；When classification, can according to behavior pattern, to sample data into
Row classification.For example, can classify to sample data according to the behavior of online and offline, user is bought to commodity point on line
For one kind, will be bought under line commodity be divided into it is another kind of.
Third walks, and carries out data cleansing to sorted sample data, obtains feature samples data.To data cleansing, packet
It includes and deletes null value, deletes wrong data etc..
4th step, according to feature samples data, the first prediction model of training.Wherein, in training prediction model, can make
With machine learning algorithm, deep learning algorithm, random forests algorithm etc. can also be used, this is suitable for the present invention.
When using random forests algorithm training pattern, parameter search can be done according to data dimension, can trained simultaneously multiple
The model of different parameters.
It by classifying to sample data, can avoid being trained unnecessary sample data, save calculation amount.
By being cleaned to sample data, the input of wrong data, assigning null data etc. can be avoided, and then prediction model can be improved
Accuracy.
In a specific embodiment provided by the invention, according to the feature samples data, training the first prediction mould
Before the step of type, can also include：The feature samples data are combined, new feature sample data is obtained；Described
According to the feature samples data, the first prediction model is trained, including：According to the new feature sample data, the first prediction of training
Model.
After obtaining feature samples data, according to feature samples data, the step of the first prediction model of training before, also
May include：Feature samples data are combined, new feature sample data is obtained.
When being combined to feature samples data, the relevance between feature can be analyzed, various features are subjected to group
It closes, does multi-ply linear transformation, and feature space is rotated, more valuable feature can be obtained.For example, we have filled very
More APP, these APP may be certain type of APP, such as ofo, and Mo Bai, little Lan bicycles etc., these are all shared bicycles,
Although being different APP, represent as being meant that, is all to there is the talent conference for the demand of riding to install；The present invention can be with
To installing ofo, Mo Bai, these features such as little Lan bicycles, which are combined, to summarize, and obtains " having the talent conference for the demand of riding to install "
New feature.Similarly, the present invention may also apply to the APP for class of managing money matters, this is all within the scope of the present invention.
By being combined to feature samples data, more new features can be obtained, these new feature collection of training are passed through
Close and obtain prediction model, accuracy, the real-time of prediction model can be improved, can significantly reduce the time, it is artificial at
This, provides correct guiding, and then greatly improves operation, advertisement delivery effect for operation, advertisement dispensing.
It in a specific embodiment provided by the invention, is combined, obtains new special to the feature samples data
After the step of levying sample data, can also include：The new feature sample data is normalized；It is described according to institute
Feature samples data are stated, the first prediction model is trained, including：According to the new feature sample data after normalized, instruction
Practice the first prediction model.
After being combined to obtain new feature sample data to feature samples data, may exist between same characteristic features
The prodigious situation of codomain span.By the way that feature is normalized, i.e., maximin standardization is done to each feature：
(x-min (x))/(max (x)-min (x)) between thus the characteristic value of these row being zoomed to 0-1, and then can be carried
High calculating speed.
For example, there are one have following two data in feature：A, user is flush of money, and residence room rate is 150000 yuan/it is flat
Rice, B, another user are to rent a house, and rent only has 200 yuan/month, and the difference between the two values is too big, in calculating below
Prediction model convergence rate can be caused extremely slow, influence working efficiency, after doing normalized to all features, calculating can be improved
Speed.
In a specific embodiment provided by the invention, after the step of training the first prediction model, it can also wrap
It includes：Obtain test data；According to the test data, the accuracy score value of first prediction model is calculated；Judge the standard
Whether exactness score value is more than default accuracy threshold value；If more than, then execute it is described according to the behavioral data, based on pre-establishing
The first prediction model, obtain the first prediction result the step of；If being not more than, re-execute described according to the feature samples
The step of data, the first prediction model of training.
When the first prediction model is two disaggregated model, when testing the first prediction model using test data, each two
The prediction result of disaggregated model can all make comparisons with the truthful data in test data, by calculating F1score come evaluation model
Quality：F1=2TP/ (2TP+FP+FN).F1 values are higher, show that model is better, accuracy is higher.Wherein, F1 values are two classification
The accuracy score value of model.TP is that authentic specimen is positive sample, and prediction result is the number of positive sample；FP is that authentic specimen is negative
Sample, prediction result are the number of positive sample；FN is that authentic specimen is negative sample, and prediction result is the number of negative sample.
When the first prediction model is more disaggregated models, accuracy score value is accuracy rate.It can be random by prediction model
Accuracy score value of the accuracy rate of test as more disaggregated models.
The present invention is suitable for any type prediction model, for example, earning power prediction model, gender prediction's model, age
Prediction model etc., but the accuracy threshold value of each model is different, this is arranged according to specific model.
When the accuracy score value of calculating is more than preset accuracy threshold value, then first prediction model can be used, it will
Behavioral data is input to the first prediction model, obtains the first prediction result；When the accuracy score value of calculating is not more than preset standard
When exactness threshold value, then show that first prediction model is unavailable, needs again according to feature samples data, training the first prediction mould
Type, until the accuracy score value of calculating is more than accuracy threshold value.
In the present invention, when being predicted using the first prediction model of test data pair, the mode pair of sampling may be used
First prediction model is predicted, in such manner, it is possible to ensure the accuracy rate of accuracy score value.
By calculating the accuracy score value of the first prediction model, the quality of prediction model, energy are judged according to accuracy score value
Enough ensure to obtain accurate prediction result when using prediction model.
In a specific embodiment provided by the invention, according to first prediction result and the behavioral data,
Before the step of training the second prediction model of optimization, further include：Based on first prediction model, the first prediction knot is obtained
The confidence level of fruit；Judge whether the confidence level is less than corresponding first threshold；If being less than, re-execute described in the basis
The step of feature samples data, the first prediction model of training；If being not less than, first prediction result is exported；And judge institute
State whether confidence level is more than corresponding second threshold；If more than then judging that first prediction result can be used as training characteristics；If
It is not more than, then judges that first prediction result not can be used as training characteristics；It is described according to first prediction result and described
Behavioral data, training the second prediction model of optimization, including：It is more than the described first pre- of the second threshold according to the confidence level
Survey result and the behavioral data, training the second prediction model of optimization.
After obtaining the first prediction result, the behavioral data of the first prediction result and user, training optimization the can be utilized
Two prediction models can improve the second prediction model in such manner, it is possible to train the second prediction model of optimization using more feature
Real-time and accuracy.
Before training optimizes the second prediction model, further include：Based on the first prediction model, the first prediction result is obtained
Confidence level.Wherein, when each prediction result exports, the confidence level of the prediction result can also be exported.Confidence level can be used for
Indicate the accuracy of each prediction result, confidence level is higher, shows that prediction result is more accurate.
When confidence level is less than first threshold, then show that the prediction result is unreliable, re -training first is needed to predict mould
Type；If being not less than, show that the result can be used, guiding can be provided for advertisement promotion；Although prediction result can make
With, but not necessarily the prediction result can be as the training sample data or optimization sample data of the second prediction model.When pre-
Survey result confidence level be more than second threshold when, then judge that the prediction result can be used as training characteristics, when no more than when, then show
The prediction result not can be used as training characteristics.
If prediction result can be as the training characteristics of the second prediction model, it is possible to according to the prediction result and row
For data, training the second prediction model of optimization.In this way, the accuracy by ensuring training data, can improve the second prediction mould
The accuracy of type.
For example, the user behavior data of acquisition does not have this feature of gender, but the first prediction model can be used for predicting
Sex character can ensure the accuracy of the prediction result of the first prediction model, so by the verification to the first prediction model
Afterwards, can utilize the first prediction model gender prediction's result training optimization the second prediction model, using the second prediction model come
It predicts hobby feature, trains the second prediction model of optimization by the way that sex character is added, hobby feature can be improved
The accuracy of prediction.
By means of the invention it is also possible to many abundant features are obtained, also, carrying with some model accuracy
The accuracy of height, other models can also be improved and then, and then being capable of the other models of real-time optimization.
More than, for a kind of optimization method of user's portrait model provided by the invention.
Second embodiment：
In above-mentioned first embodiment, a kind of optimization method of user's portrait model is provided, it is real in conjunction with above-mentioned first
Example is applied, second embodiment of the invention provides a kind of computer readable storage medium, is stored thereon with computer program, the program quilt
Processor realizes a kind of optimization method for user portrait model that above-mentioned first embodiment provides when executing.
3rd embodiment：
In conjunction with a kind of optimization method for user portrait model that first embodiment provides, the present invention also provides a kind of user pictures
As the optimization equipment of model, including：Memory, processor and storage are on a memory and the computer that can run on a processor
Program, the processor realize a kind of optimization side for user portrait model that above-mentioned first embodiment provides when executing described program
Method.Fig. 2 shows a kind of hardware architecture diagrams of the optimization equipment of user's portrait model provided in an embodiment of the present invention.
Specifically, above-mentioned processor 201 may include central processing unit (CPU) or specific integrated circuit
(Application Specific Integrated Circuit, ASIC), or may be configured to implement implementation of the present invention
One or more integrated circuits of example.
Memory 202 may include the mass storage for data or instruction.For example unrestricted, memory
202 may include hard disk drive (Hard Disk Drive, HDD), floppy disk, flash memory, CD, magneto-optic disk, tape or logical
With the combination of universal serial bus (Universal Serial Bus, USB) driver or two or more the above.It is closing
In the case of suitable, memory 202 may include the medium of removable or non-removable (or fixed).In a suitable case, it stores
Device 202 can be inside or outside data processing equipment.In a particular embodiment, memory 202 is nonvolatile solid state storage
Device.In a particular embodiment, memory 202 includes read-only memory (ROM).In a suitable case, which can be mask
The ROM of programming, programming ROM (PROM), erasable PROM (EPROM), electric erasable PROM (EEPROM), electrically-alterable ROM
(EAROM) or the combination of flash memory or two or more the above.
Processor 201 is by reading and executing the computer program instructions stored in memory 202, to realize above-mentioned implementation
The optimization method of any one user portrait model in example.
In one example, the optimization equipment of user's portrait model may also include communication interface 203 and bus 210.Wherein,
As shown in Fig. 2, processor 201, memory 202, communication interface 203 are connected by bus 210 and complete mutual communication.
Communication interface 203 is mainly used for realizing in the embodiment of the present invention between each module, device, unit and/or equipment
Communication.
Bus 210 includes hardware, software or both, and the component of the optimization equipment of user's portrait model is coupled to each other one
It rises.For example unrestricted, bus may include accelerated graphics port (AGP) or other graphics bus, enhancing industrial standard frame
Structure (EISA) bus, front side bus (FSB), super transmission (HT) interconnection, Industry Standard Architecture (ISA) bus, infinite bandwidth interconnection,
Low pin count (LPC) bus, memory bus, micro- channel architecture (MCA) bus, peripheral component interconnection (PCI) bus, PCI-
Express (PCI-X) bus, Serial Advanced Technology Attachment (SATA) bus, Video Electronics Standards Association part (VLB) bus or
The combination of other suitable buses or two or more the above.In a suitable case, bus 210 may include one
Or multiple buses.Although specific bus has been described and illustrated in the embodiment of the present invention, the present invention considers any suitable bus
Or interconnection.
It should be clear that the invention is not limited in specific configuration described above and shown in figure and processing.
For brevity, it is omitted here the detailed description to known method.In the above-described embodiments, several tools have been described and illustrated
The step of body, is as example.But procedure of the invention is not limited to described and illustrated specific steps, this field
Technical staff can be variously modified, modification and addition after the spirit for understanding the present invention, or suitable between changing the step
Sequence.
Functional block shown in above structure diagram can be implemented as hardware, software, firmware or combination thereof.When
When realizing in hardware, electronic circuit, application-specific integrated circuit (ASIC), firmware appropriate, plug-in unit, function may, for example, be
Card etc..When being realized with software mode, element of the invention is used to execute the program or code segment of required task.Journey
Sequence either code segment can be stored in machine readable media or the data-signal by being carried in carrier wave in transmission medium or
Person's communication links are sent." machine readable media " may include any medium for capableing of storage or transmission information.It is machine readable
The example of medium include electronic circuit, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disk, CD-ROM,
CD, hard disk, fiber medium, radio frequency (RF) link, etc..Code segment can be via the calculating of internet, Intranet etc.
Machine network is downloaded.
Finally it should be noted that：The above embodiments are only used to illustrate the technical solution of the present invention., rather than its limitations；To the greatest extent
Present invention has been described in detail with reference to the aforementioned embodiments for pipe, it will be understood by those of ordinary skill in the art that：Its according to
So can with technical scheme described in the above embodiments is modified, either to which part or all technical features into
Row equivalent replacement；And these modifications or replacements, various embodiments of the present invention technology that it does not separate the essence of the corresponding technical solution
The range of scheme should all cover in the claim of the present invention and the range of specification.
Claims (9)
- The optimization method of model 1. a kind of user draws a portrait, which is characterized in that including：Obtain user behavior data；The first prediction result is obtained based on the first prediction model pre-established according to the behavioral data；According to first prediction result and the behavioral data, training the second prediction model of optimization.
- 2. according to the method described in claim 1, it is characterized in that, according to first behavioral data, based on pre-establishing The first prediction model, obtain the first prediction result the step of before, further include：Obtain sample data；Classify to the sample data；Data cleansing is carried out to the sorted sample data, obtains the feature samples data of the sample data；According to the feature samples data, the first prediction model of training.
- 3. according to the method described in claim 2, it is characterized in that, described classify to the sample data, including：According to behavior pattern, classify to the sample data.
- 4. according to the method described in claim 2, it is characterized in that, according to the feature samples data, the first prediction of training Before the step of model, further include：The feature samples data are combined, new feature sample data is obtained；It is described to train the first prediction model according to the feature samples data, including：According to the new feature sample data, the first prediction model of training.
- 5. according to the method described in claim 4, it is characterized in that, being combined to the feature samples data, acquisition is new After the step of feature samples data, further include：The new feature sample data is normalized；It is described to train the first prediction model according to the feature samples data, including：According to the new feature sample data after normalized, the first prediction model of training.
- 6. according to the method described in claim 2, it is characterized in that, after the step of training the first prediction model, further include：Obtain test data；According to the test data, the accuracy score value of first prediction model is calculated；Judge whether the accuracy score value is more than default accuracy threshold value；If more than then executing described according to the behavioral data, based on the first prediction model pre-established, obtain the first prediction As a result the step of；If being not more than, re-execute it is described according to the feature samples data, the step of the first prediction model of training.
- 7. according to the method described in claim 6, it is characterized in that, according to first prediction result and the behavior number According to, training optimization the second prediction model the step of before, further include：Based on first prediction model, the confidence level of first prediction result is obtained；Judge whether the confidence level is less than corresponding first threshold；If being less than, re-execute described according to the feature sample The step of notebook data, the first prediction model of training；If being not less than, first prediction result is exported；And judge whether the confidence level is more than corresponding second threshold；If more than then judging that first prediction result can be used as training characteristics；If being not more than, the first prediction knot is judged Fruit not can be used as training characteristics；It is described that according to first prediction result and the behavioral data, training optimizes the second prediction model, including：It is more than first prediction result of the second threshold and the behavioral data, training optimization the according to the confidence level Two prediction models.
- 8. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the program is held by processor The method described in one of claim 1-7 is realized when row.
- The optimization equipment of model 9. a kind of user draws a portrait, including：Memory, processor and storage on a memory and can handled The computer program run on device, which is characterized in that the processor realizes one of claim 1-7 institutes when executing described program The method stated.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810035915.4A CN108280542B (en) | 2018-01-15 | 2018-01-15 | User portrait model optimization method, medium and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810035915.4A CN108280542B (en) | 2018-01-15 | 2018-01-15 | User portrait model optimization method, medium and equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108280542A true CN108280542A (en) | 2018-07-13 |
| CN108280542B CN108280542B (en) | 2021-05-11 |
Family
ID=62803631
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810035915.4A Active CN108280542B (en) | 2018-01-15 | 2018-01-15 | User portrait model optimization method, medium and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108280542B (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109447694A (en) * | 2018-10-11 | 2019-03-08 | 上海瀚之友信息技术服务有限公司 | A kind of user feature analysis method and its system |
| CN109933698A (en) * | 2019-02-27 | 2019-06-25 | 腾讯科技（深圳）有限公司 | A kind of the source method of calibration and device of user's portrait |
| CN110083517A (en) * | 2019-04-29 | 2019-08-02 | 秒针信息技术有限公司 | A kind of optimization method and device of user's portrait confidence level |
| CN110134722A (en) * | 2019-05-22 | 2019-08-16 | 北京小度信息科技有限公司 | Target user determines method, apparatus, equipment and storage medium |
| CN110634044A (en) * | 2019-08-09 | 2019-12-31 | 万翼科技有限公司 | Park user behavior analysis method and device, computer equipment and storage medium |
| WO2020042584A1 (en) * | 2018-08-27 | 2020-03-05 | 平安科技（深圳）有限公司 | Method and system for training underperforming personnel, computer device, and storage medium |
| CN111091489A (en) * | 2019-11-01 | 2020-05-01 | 平安科技（深圳）有限公司 | Picture optimization method and device, electronic equipment and storage medium |
| CN112133431A (en) * | 2020-08-27 | 2020-12-25 | 绿瘦健康产业集团有限公司 | Health information message pushing method, device, medium and terminal equipment |
| CN112182333A (en) * | 2020-09-25 | 2021-01-05 | 山东亿云信息技术有限公司 | Talent space-time big data processing method and system based on random forest |
| CN112967424A (en) * | 2021-02-02 | 2021-06-15 | 广州橙行智动汽车科技有限公司 | Simulation method and device of vehicle-mounted Bluetooth key |
| CN113538029A (en) * | 2020-04-22 | 2021-10-22 | 中国移动通信集团上海有限公司 | User behavior data prediction method, device, equipment and medium |
| TWI777154B (en) * | 2020-04-23 | 2022-09-11 | 和碩聯合科技股份有限公司 | Electronic device and method for turnover rate prediction |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101140623A (en) * | 2007-09-29 | 2008-03-12 | 华为技术有限公司 | Video frequency objects recognition method and system based on supporting vectors machine |
| US20090259534A1 (en) * | 2008-04-11 | 2009-10-15 | Microsoft Corporation | Internal business arbitrage |
| CN101739555A (en) * | 2009-12-01 | 2010-06-16 | 北京中星微电子有限公司 | Method and system for detecting false face, and method and system for training false face model |
| WO2012075600A1 (en) * | 2010-12-06 | 2012-06-14 | Yahoo! Inc. | System and method for list ranking and ads placement using interaction features |
| CN102567391A (en) * | 2010-12-20 | 2012-07-11 | 中国移动通信集团广东有限公司 | Method and device for building classification forecasting mixed model |
| CN103502899A (en) * | 2011-01-26 | 2014-01-08 | 谷歌公司 | Dynamic predictive modeling platform |
| CN104464291A (en) * | 2014-12-08 | 2015-03-25 | 杭州智诚惠通科技有限公司 | Traffic flow predicting method and system |
| CN104810017A (en) * | 2015-04-08 | 2015-07-29 | 广东外语外贸大学 | Semantic analysis-based oral language evaluating method and system |
| CN106776925A (en) * | 2016-11-30 | 2017-05-31 | 腾云天宇科技（北京）有限公司 | A kind of Forecasting Methodology of mobile terminal user's sex, server and system |
| CN107330445A (en) * | 2017-05-31 | 2017-11-07 | 北京京东尚科信息技术有限公司 | The Forecasting Methodology and device of user property |
-
2018
- 2018-01-15 CN CN201810035915.4A patent/CN108280542B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101140623A (en) * | 2007-09-29 | 2008-03-12 | 华为技术有限公司 | Video frequency objects recognition method and system based on supporting vectors machine |
| US20090259534A1 (en) * | 2008-04-11 | 2009-10-15 | Microsoft Corporation | Internal business arbitrage |
| CN101739555A (en) * | 2009-12-01 | 2010-06-16 | 北京中星微电子有限公司 | Method and system for detecting false face, and method and system for training false face model |
| WO2012075600A1 (en) * | 2010-12-06 | 2012-06-14 | Yahoo! Inc. | System and method for list ranking and ads placement using interaction features |
| CN102567391A (en) * | 2010-12-20 | 2012-07-11 | 中国移动通信集团广东有限公司 | Method and device for building classification forecasting mixed model |
| CN103502899A (en) * | 2011-01-26 | 2014-01-08 | 谷歌公司 | Dynamic predictive modeling platform |
| CN104464291A (en) * | 2014-12-08 | 2015-03-25 | 杭州智诚惠通科技有限公司 | Traffic flow predicting method and system |
| CN104810017A (en) * | 2015-04-08 | 2015-07-29 | 广东外语外贸大学 | Semantic analysis-based oral language evaluating method and system |
| CN106776925A (en) * | 2016-11-30 | 2017-05-31 | 腾云天宇科技（北京）有限公司 | A kind of Forecasting Methodology of mobile terminal user's sex, server and system |
| CN107330445A (en) * | 2017-05-31 | 2017-11-07 | 北京京东尚科信息技术有限公司 | The Forecasting Methodology and device of user property |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020042584A1 (en) * | 2018-08-27 | 2020-03-05 | 平安科技（深圳）有限公司 | Method and system for training underperforming personnel, computer device, and storage medium |
| CN109447694B (en) * | 2018-10-11 | 2022-04-12 | 上海瀚之友信息技术服务有限公司 | User characteristic analysis method and system |
| CN109447694A (en) * | 2018-10-11 | 2019-03-08 | 上海瀚之友信息技术服务有限公司 | A kind of user feature analysis method and its system |
| CN109933698A (en) * | 2019-02-27 | 2019-06-25 | 腾讯科技（深圳）有限公司 | A kind of the source method of calibration and device of user's portrait |
| CN110083517A (en) * | 2019-04-29 | 2019-08-02 | 秒针信息技术有限公司 | A kind of optimization method and device of user's portrait confidence level |
| CN110083517B (en) * | 2019-04-29 | 2023-08-25 | 秒针信息技术有限公司 | User image confidence optimization method and device |
| CN110134722A (en) * | 2019-05-22 | 2019-08-16 | 北京小度信息科技有限公司 | Target user determines method, apparatus, equipment and storage medium |
| CN110634044A (en) * | 2019-08-09 | 2019-12-31 | 万翼科技有限公司 | Park user behavior analysis method and device, computer equipment and storage medium |
| CN111091489A (en) * | 2019-11-01 | 2020-05-01 | 平安科技（深圳）有限公司 | Picture optimization method and device, electronic equipment and storage medium |
| CN111091489B (en) * | 2019-11-01 | 2024-05-07 | 平安科技（深圳）有限公司 | Picture optimization method and device, electronic equipment and storage medium |
| CN113538029A (en) * | 2020-04-22 | 2021-10-22 | 中国移动通信集团上海有限公司 | User behavior data prediction method, device, equipment and medium |
| TWI777154B (en) * | 2020-04-23 | 2022-09-11 | 和碩聯合科技股份有限公司 | Electronic device and method for turnover rate prediction |
| CN112133431A (en) * | 2020-08-27 | 2020-12-25 | 绿瘦健康产业集团有限公司 | Health information message pushing method, device, medium and terminal equipment |
| CN112182333A (en) * | 2020-09-25 | 2021-01-05 | 山东亿云信息技术有限公司 | Talent space-time big data processing method and system based on random forest |
| CN112967424A (en) * | 2021-02-02 | 2021-06-15 | 广州橙行智动汽车科技有限公司 | Simulation method and device of vehicle-mounted Bluetooth key |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108280542B (en) | 2021-05-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108280542A (en) | A kind of optimization method, medium and the equipment of user's portrait model | |
| CN110147456B (en) | Image classification method and device, readable storage medium and terminal equipment | |
| CN107491432B (en) | Low-quality article identification method and device based on artificial intelligence, equipment and medium | |
| CN109344257B (en) | Text emotion recognition method and device, electronic equipment and storage medium | |
| CN108985489B (en) | Risk prediction method, risk prediction device and terminal equipment | |
| CN106651057A (en) | Mobile terminal user age prediction method based on installation package sequence table | |
| CN111028016A (en) | Sales data prediction method and device and related equipment | |
| CN107657056A (en) | Method and apparatus based on artificial intelligence displaying comment information | |
| CN106372057A (en) | Content auditing method and apparatus | |
| CN108734296A (en) | Optimize method, apparatus, electronic equipment and the medium of the training data of supervised learning | |
| CN111582315B (en) | Sample data processing method and device and electronic equipment | |
| CN111343147A (en) | Network attack detection device and method based on deep learning | |
| CN107229731A (en) | Method and apparatus for grouped data | |
| CN109961311A (en) | Lead referral method, apparatus calculates equipment and storage medium | |
| CN111582645B (en) | APP risk assessment method and device based on factoring machine and electronic equipment | |
| CN102955950A (en) | Device for online training classifier and method for online training classifier | |
| CN111861521A (en) | Data processing method and device, computer readable medium and electronic equipment | |
| CN112328869A (en) | User loan willingness prediction method and device and computer system | |
| CN111210332A (en) | Method and device for generating post-loan management strategy and electronic equipment | |
| CN113919432A (en) | Classification model construction method, data classification method and device | |
| KR102152081B1 (en) | Valuation method based on deep-learning and apparatus thereof | |
| CN112685374B (en) | Log classification method and device and electronic equipment | |
| CN111385659B (en) | Video recommendation method, device, equipment and storage medium | |
| CN107909414A (en) | The anti-cheat method and device of application program | |
| CN111666748A (en) | Construction method of automatic classifier and method for recognizing decision from software development text product |
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 | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: Optimization method, medium and device of a user portrait model Effective date of registration: 20221031 Granted publication date: 20210511 Pledgee: Shenzhen small and medium sized small loan Co.,Ltd. Pledgor: SHENZHEN HEXUN HUAGU INFORMATION TECHNOLOGY CO.,LTD. Registration number: Y2022440020221 |