CN109102159A - Passenger's rating model generation method, device, computer equipment and storage medium - Google Patents
Passenger's rating model generation method, device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN109102159A CN109102159A CN201810788417.7A CN201810788417A CN109102159A CN 109102159 A CN109102159 A CN 109102159A CN 201810788417 A CN201810788417 A CN 201810788417A CN 109102159 A CN109102159 A CN 109102159A
- Authority
- CN
- China
- Prior art keywords
- rating model
- standard
- place
- passenger
- record
- 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
- 238000000034 method Methods 0.000 title claims abstract description 35
- 238000012549 training Methods 0.000 claims description 48
- 238000012360 testing method Methods 0.000 claims description 34
- 230000000295 complement effect Effects 0.000 claims description 33
- 238000012795 verification Methods 0.000 claims description 28
- 238000004590 computer program Methods 0.000 claims description 25
- 238000004364 calculation method Methods 0.000 claims description 8
- 239000000284 extract Substances 0.000 claims description 5
- 239000013589 supplement Substances 0.000 claims description 5
- 238000003066 decision tree Methods 0.000 description 14
- 241000208340 Araliaceae Species 0.000 description 4
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 4
- 235000003140 Panax quinquefolius Nutrition 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 235000008434 ginseng Nutrition 0.000 description 4
- 238000007689 inspection Methods 0.000 description 4
- 230000003068 static effect Effects 0.000 description 4
- 239000012141 concentrate Substances 0.000 description 3
- 238000000605 extraction Methods 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 229910052739 hydrogen Inorganic materials 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- 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
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Economics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- General Business, Economics & Management (AREA)
- Educational Administration (AREA)
- Physics & Mathematics (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Game Theory and Decision Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
This application involves a kind of passenger's rating model generation method, device, computer equipment and storage mediums.It reaches a standard record the described method includes: obtaining history, the history record that reaches a standard carries risk passenger label or average traveler label；The history is reached a standard to record and is grouped according to place of reaching a standard；It reaches a standard to record to the history after grouping and is trained to obtain corresponding place rating model respectively；It is combined obtained place rating model to obtain passenger's rating model.Grading accuracy can be improved using this method.
Description
Technical field
This application involves field of computer technology, more particularly to a kind of passenger's rating model generation method, device, calculating
Machine equipment and storage medium.
Background technique
The entry and exit place such as airport, port daily all can a large amount of passengers reach a standard, wherein be no lack of it is some smuggle, steal into another country etc. it is illegal
Molecule.There is the current record of a large amount of personnel in security protection scene at present, there are small part personnel that there is high risk in current crowd.
However, security personnel only can by rule of thumb spot-check the passenger that reaches a standard at present, or with simple rule to a large amount of
Data carry out preliminary screening, and both methods can not accurately grade to the passenger that reaches a standard.
Summary of the invention
Based on this, it is necessary in view of the above technical problems, provide a kind of passenger's grading mould that can be improved grading accuracy
Type generation method, device, computer equipment and storage medium.
A kind of passenger's rating model generation method, which comprises
It obtains history to reach a standard record, the history record that reaches a standard carries risk passenger label or average traveler label；
The history is reached a standard to record and is grouped according to place of reaching a standard；
It reaches a standard to record to the history after grouping and is trained to obtain corresponding place rating model respectively；
It is combined obtained place rating model to obtain passenger's rating model.
It is described in one of the embodiments, to be combined obtained place rating model to obtain passenger's grading mould
Type, comprising:
The weight distribution for the place rating model for receiving input instructs；
It is instructed to obtain the weight of the place rating model according to the weight distribution；
Passenger's grading is obtained according to the place rating model and corresponding weight calculation according to average weighted mode
Model.
The record that reaches a standard of the history after described pair of grouping is trained to obtain and comments on correspondingly in one of the embodiments,
Grade model, comprising:
Current business rule is obtained, and inquires complementary features parameter corresponding with the current business rule；
Respectively according to described first after grouping reach a standard record in primitive character parameter calculate complementary features parameter；
The complementary features parameter and primitive character parameter are trained to obtain corresponding place rating model.
The record that reaches a standard of the history after described pair of grouping is trained to obtain and comments on correspondingly in one of the embodiments,
Grade model, comprising:
The history after the grouping record that reaches a standard is divided into training set data and test set data；
Fisrt feature parameter is extracted from the training set data, and attribute gain is carried out according to the fisrt feature parameter and is commented
Estimate, and target signature is extracted from the fisrt feature parameter according to attribute gain assessment result；
According to extracted target signature and the corresponding passenger's label of the corresponding training set data to the training
Collection data are classified to obtain initial rating model, and calculate the scoring etc. of each class node in the initial rating model
Grade；
Second feature parameter corresponding with selected target signature is extracted from the test set data；
By the second feature parameter and the corresponding passenger's label of the test set data to the initial grading mould
The grading system of each node is verified to obtain the first verification result in type；
The initial rating model is adjusted to obtain place rating model according to first verification result.
In one of the embodiments, the method also includes:
When reach history reach a standard record renewal time when, the history for loading update reaches a standard record；
It reaches a standard from the history of update and extracts third feature parameter corresponding with the place rating model in recording；
It is reached a standard according to the history of the third feature parameter and update and records corresponding passenger's label to the place
The grading system of each node is verified to obtain the second verification result in rating model；
The place rating model is optimized according to second verification result.
A kind of passenger's rating model generating means, described device include:
The history record that reaches a standard obtains module, reaches a standard record for obtaining history, and the history record carrying that reaches a standard is risky
Passenger's label or average traveler label；
Grouping module is grouped for the history to reach a standard to record according to place of reaching a standard；
Place rating model generation module, for the history after grouping reach a standard record be trained to obtain respectively it is corresponding
Place rating model；
Passenger's rating model generation module obtains passenger's grading mould for obtained place rating model to be combined
Type.
Passenger's rating model generation module includes: in one of the embodiments,
Receiving unit, the weight distribution for receiving input for the place rating model instruct；
Weight Acquisition unit obtains the weight of the place rating model for instructing according to the weight distribution；
First computing unit is used for according to average weighted mode according to the place rating model and corresponding weight
Passenger's rating model is calculated.
The place rating model generation module includes: in one of the embodiments,
Business rule acquiring unit for obtaining current business rule, and is inquired corresponding with the current business rule
Complementary features parameter；
Second computing unit, for respectively according to described first after grouping reach a standard record in primitive character parameter calculate
Complementary features parameter；
Place rating model generation unit, for being trained to the complementary features parameter and primitive character parameter
To corresponding place rating model.
A kind of computer equipment, including memory and processor, the memory are stored with computer program, the processing
The step of device realizes any of the above-described the method when executing the computer program.
A kind of computer readable storage medium, is stored thereon with computer program, and the computer program is held by processor
The step of method described in any of the above embodiments is realized when row.
Above-mentioned passenger's rating model generation method, device, computer equipment and storage medium, to history reach a standard record according to
Place of reaching a standard is grouped, and generates corresponding place rating model respectively according to the difference in place of reaching a standard, and finally comments on ground
Grade model is combined to have obtained passenger's rating model, so that passenger's rating model coverage is wide, so as to improve passenger
The accuracy of grading.
Detailed description of the invention
Fig. 1 is the application scenario diagram of passenger's rating model generation method in one embodiment；
Fig. 2 is the flow diagram of passenger's rating model generation method in one embodiment；
Fig. 3 is the structural block diagram of passenger's rating model generating means in one embodiment；
Fig. 4 is the internal structure chart of computer equipment in one embodiment.
Specific embodiment
It is with reference to the accompanying drawings and embodiments, right in order to which the objects, technical solutions and advantages of the application are more clearly understood
The application is further elaborated.It should be appreciated that specific embodiment described herein is only used to explain the application, not
For limiting the application.
Passenger's rating model generation method provided by the present application, can be applied in application environment as shown in Figure 1.Its
In, server 102 is communicated by network and database 104.Server 102 can read history mistake from database 104
Close record, which reaches a standard and carry risky trip's label or average traveler label in record, and server first reaches a standard history note
Record is grouped according to place of reaching a standard, and then the history after grouping reaches a standard to record is trained to obtain respectively to comment on correspondingly
Grade model；It finally is combined obtained place rating model to obtain passenger's rating model, so that passenger's rating model relates to
And range is wide, so as to improve the accuracy of passenger's grading.Wherein, server 102 can be with independent server either
The server cluster of multiple servers composition is realized.
In one embodiment, it as shown in Fig. 2, providing a kind of passenger's rating model generation method, applies in this way
It is illustrated for server in Fig. 1, comprising the following steps:
S202: obtaining history and reach a standard record, and the history record that reaches a standard carries risk passenger label or average traveler label.
Specifically, history reaches a standard record generated record when being previous passenger's clearance, contains in record if this reaches a standard
Dry elementary field, including name, age, gender and clearance time etc., which may include not examined passenger
Reach a standard record and the record that reaches a standard of examined passenger, it is examined passenger that the reaching a standard of examined passenger, which records corresponding,
Whether be risk passenger record, wherein for convenience, not examined passenger and examined and inspection result are common
The corresponding record that reaches a standard of passenger carry average traveler label, examined and inspection result, which is that risk passenger is corresponding, to reach a standard
Record carries risk passenger label.
Wherein server can obtain whole history from database and reach a standard record, or for convenience, server sheet
Ground can also be cached with partial history and reach a standard record, to facilitate model training.
S204: history is reached a standard to record is grouped according to place of reaching a standard.
Specifically, it reaches a standard record due to storing a large amount of history in database, it for convenience can be by history mistake
Record is closed to be grouped according to place of reaching a standard.Such as server detection history can reach a standard the Locality field that reaches a standard of record first,
Then the acquired history record that reaches a standard is grouped by the Locality field that reaches a standard.
S206: reaching a standard to record to the history after grouping is trained to obtain corresponding place rating model respectively.
It specifically, can be to the history mistake after grouping after the record that reaches a standard to history is grouped according to place of just reaching a standard
Record is closed to be trained to obtain the corresponding place rating model in place of reaching a standard respectively.And each place rating model is to use
What decision-tree model obtained, and passenger can be predicted in the risk class in place that reaches a standard.
S208: it is combined obtained place rating model to obtain passenger's rating model.
Specifically, for server after training obtains place rating model, the place rating model obtained to training carries out group
Conjunction obtains passenger's rating model, such as can distribute different weights to different place rating models, then according to weight with
And different place rating models obtains passenger's rating model, wherein the weight distributed can be according to each place grading mould
What the size of the ratio of the historical record that reaches a standard in the corresponding historical record that reaches a standard of type with risk passenger label was configured, example
Such as when the corresponding aforementioned proportion in a certain place of reaching a standard is larger, then its weight becomes larger.Optionally, server is available to each
Reach a standard the corresponding ratio in place, and then those ratios are normalized as last weight.
Above-mentioned passenger's rating model generation method, reaching a standard to record to history is grouped according to place of reaching a standard, and according to mistake
The difference for closing place generates corresponding place rating model respectively, finally is combined the place rating model to have obtained passenger and comment
Grade model, so that passenger's rating model coverage is wide, so as to improve the accuracy of passenger's grading.
It is combined obtained place rating model to obtain passenger's rating model in one of the embodiments, it can
To include: the weight distribution instruction for place rating model for receiving input；It is instructed to obtain place grading according to weight distribution
The weight of model；Passenger's grading mould is obtained according to average weighted mode base area point rating model and corresponding weight calculation
Type.
The record that reaches a standard of the history after grouping is trained to obtain corresponding place and graded mould in one of the embodiments,
Type may include: to obtain current business rule, and inquire complementary features parameter corresponding with current business rule；Basis respectively
First after grouping reach a standard record in primitive character parameter calculate complementary features parameter；To complementary features parameter and original spy
Sign parameter is trained to obtain corresponding place rating model.
Specifically, by obtained place rating model be combined to obtain passenger's rating model can be by user into
What row was intervened, such as server can export each place rating model, and show the corresponding mistake of each place rating model
Place is closed, user can distribute corresponding weight to each place rating model according to actual needs, for example, if reaching a standard place
It is the first place, then the weight of the corresponding place rating model in the first place can be arranged in correspondence with larger by user, and will be with
The weight of the place rating model in associated second place in the first place be arranged in correspondence with it is larger, and by other place rating models
Weight be arranged in correspondence with it is less.It wherein can be that there are business relations with the first place with associated second place in the first place
The second place.
Server is after the corresponding weight of place rating model for receiving user's input, then according to average weighted mode
Base area point rating model and corresponding weight calculation obtain passenger's rating model.Such as A (passenger's rating model)=(A1*a1+
A2*a2+A3*a3+ ...+AN*an)/N, wherein A1, A2 ... AN refer to that model, a1, a2 ... an refer to weight.
Specifically, available to current business rule before above-mentioned place rating model generates, and inquire current industry
The corresponding complementary features parameter of business rule, so as to be calculated according to the primitive character parameter in record that reaches a standard of first after grouping
Complementary features parameter.Wherein current business rule includes current feature, checks category feature, the information that reaches a standard category feature, frequency static
Feature, frequency dynamic feature；To which complementary features parameter generated may include: colleague's class complementary features parameter: 30, such as
" before 15 days, number of going together in 7 days ", " before 15 days, everyone average number of going together in 30 days " etc..Check class complementary features ginseng
Number: 20, such as " before 30 days, hits in 30 days ", " before 30 days, the interval number of days of the last clearance and examination time "
Deng.Reach a standard info class complementary features parameter: 23, such as " before 15 days, the last clearance time ", " before 30 days, in 30 days
Critical point number " etc..Frequency class static state complementary features parameter: 14, such as " before 15 days, clearance number in 30 days ", " before 30 days, 7
Clearance number of days in it " etc..Frequency class dynamic complementary features parameter: 34, such as " before 15 days, in 90 days for the first time and the 5th time
Minimum time interval ", " before 30 days, first time and the 5th minor tick time average in 30 days " etc..
Optionally, it when the current business rule that server is got is same line discipline, is then got by closing in record
Colleague number of the different passengers in preset number of days, will be to colleague's number feature as complementary features parameter.Similarly, reach a standard letter
It ceases category feature, frequency class static nature and frequency dynamic category feature also in the same way, is generated pair according to business rule
The logic answered.
Complementary features parameter and primitive character parameter are trained to obtain by server after generating complementary features parameter
Complementary features parameter and primitive character parameter can be combined and are trained by place rating model, optionally, server
It obtains several place rating models, and the verifying history record that reaches a standard is input in several place rating models generated and obtains
The risk class of passenger, and be compared with the history risk class in record that reaches a standard, obtain several places grading generated
The corresponding success rate of model, obtains the risk class of passenger in the success rate=place rating model and history reaches a standard in record
Reach a standard quantity/verifying history of record of the identical history of risk class reaches a standard record, and server is selected to prominent place
Place rating model of the rating model as the place of reaching a standard, and the corresponding feature of the maximum place rating model of success rate is made
For training characteristics.
In above-described embodiment, before training place rating model, multiple supplements are generated according to the history record that reaches a standard first
Characteristic parameter, so that the place rating model generated is more accurate.And user can according to need to configure each place
The weight of rating model, so that passenger's rating model coverage generated is wide, so as to improve passenger's grading
Accuracy.
The record that reaches a standard of the history after grouping is trained to obtain corresponding place and graded mould in one of the embodiments,
Type may include: that the record that reaches a standard of the history after grouping is divided into training set data and test set data；From training set data
Fisrt feature parameter is extracted, attribute gain assessment is carried out according to fisrt feature parameter, and according to attribute gain assessment result from the
Target signature is extracted in one characteristic parameter；According to extracted target signature and the corresponding passenger's mark of corresponding training set data
Label classify training set data to obtain initial rating model, and calculate the scoring of each class node in initial rating model
Grade；Second feature parameter corresponding with selected target signature is extracted from test set data；Pass through second feature parameter
And the corresponding passenger's label of test set data verifies the grading system of each node in initial rating model to obtain
One verification result；Initial rating model is adjusted according to the first verification result to obtain place rating model.
Above-mentioned passenger's rating model generation method can also include: when arrival history reaches a standard in one of the embodiments,
When recording renewal time, the history for loading update reaches a standard record；It reaches a standard and is extracted in recording and grading mould in place from the history of update
The corresponding third feature parameter of type；It is reached a standard according to the history of third feature parameter and update and records corresponding passenger's label to place
The grading system of each node is verified to obtain the second verification result in rating model；Put grading over the ground according to the second verification result
Model optimizes.
Specifically, it is illustrated so that A reaches a standard place as an example in the present embodiment, wherein A is reached a standard the corresponding history in place
The record that reaches a standard is divided into training set data and test set data, and fisrt feature parameter and the first mesh are extracted from training set data
Mark classification (corresponding passenger's grading system)；Characteristic information gain assessment is carried out according to fisrt feature parameter, is commented according to characteristic information
Estimate result and carry out feature selecting, i.e. selection target feature, then extracted target signature and corresponding training set data pair
The passenger's label answered classifies training set data to obtain initial rating model, and calculates each in initial decision tree assessment models
The grading system of node；Second feature parameter and the second target category are extracted from test set data；Joined according to second feature
Several and the second target category verifies the grading system of each node in initial decision assessment models；According to the first verification result
Adjustment is optimized to the decision tree structure in initial rating model and generates place rating model.In the present embodiment first
Characteristic parameter and second feature parameter are and the complementary features parameter and primitive character parameter that are previously mentioned in above-described embodiment, target
Classification is divided into multiclass, the i.e. grading system of passenger, and the grading system of the sample data is to have known in advance.
Decision tree be it is a kind of be made of node and directed edge, the tree structure for classifying to example.Node
There are two types of types: internal node and leaf node.Wherein, internal node indicates the test condition of feature or attribute, leaf node
Presentation class.It is using the specific method that decision-tree model is classified: since root node, a certain feature of example is carried out
Test, according to test result by example allocation to its child node.Leaf node is likely to be breached along the branch or reaches another
When internal node, then gone down using new test condition recurrence execution, until arriving at a leaf node.When arrival leaf node
When, then obtain final classification result.
Decision-tree model uses ID3 algorithm in the present embodiment, is more better than big decision based on small-sized decision tree
The principle of tree is assessed and is selected according to information gain feature, selects the maximum feature of information gain to build as Rule of judgment every time
Vertical child node.Information gain indicates to learn the information of feature X and make the degree of the uncertain reduction of the information of class Y.Feature A
To the information gain g (D, A) of training dataset D, it is defined as the experience of D under the empirical entropy H (D) and feature A specified criteria of set D
The difference of conditional entropy H (D | A), i.e.,
G (D, A)=H (D)-H (D | A) (I)
Wherein, g (D, A) is characterized A to the information gain of training dataset D, and H (D) is the empirical entropy of training dataset D, H
(D | A) A is characterized to the empirical condition entropy of data set D.
Feature selection approach according to information gain criterion is: calculating each of which feature to training dataset (or subset)
Information gain selects the maximum feature of information gain.Calculate information gain algorithm it is as follows: its input be training dataset D and
Feature A, output are characterized A to the information gain g (D, A) of training dataset D.
Firstly, calculating the empirical entropy H (D) of data set D:
Wherein, CkFor the corresponding sample size of first object classification, K is the categorical measure of first object classification, in this reality
It applies in example, first object classification is divided into each grading system of passenger.
Secondly, calculating feature A to the empirical condition entropy H of data set D (D | A):
Wherein, value (A) is all value set of feature A, and i is a value of feature A, DiIt is training dataset D
Middle feature A value is the sample set of i, | Di| indicate that value is the sample size of the sample set of i, | D | it indicates to carry out sample
The total quantity of sample before set divides, if the corresponding feature A of sex character parameter all values are male and female, as male can use
0 indicates, female can be indicated with 1, and value (A) is (0,1).
Third calculates information gain:
G (D, A)=H (D)-H (D | A) (1)
Server extracts second feature parameter and the second target category from each sample of test set data one by one.Its
In, second feature parameter is identical as the above-mentioned classification of fisrt feature parameter, it is alternatively possible to be selected information gain most
Big feature, i.e., above-mentioned target signature, details are not described herein.Second target category is safety inspection resulting class, the second mesh
Mark classification is passenger's grading system.
It, can be with if the deviation of verification result is excessive through test set data when being verified to preset rules model
Selected characteristic parameter is adjusted, such as statistic is adjusted, rebuilds the decision tree of preset rules model
And carry out verifying until the first verification result in error range, can also be since root node to the feature selecting of branch node
It is adjusted, decision tree is optimized, it, can be using modes such as the data volumes for increasing training set, until optimization in adjustment
The first verification result of decision tree can be in error range.
And getting the maximum feature of differentiation degree (field) according to characteristic information assessment result may include calculating first
The information gain of the corresponding each characteristic parameter of characteristic parameter；The maximum feature of information gain is chosen as Rule of judgment and establishes son section
Point；Be divided into subset data according to the corresponding training set data of child node, subset data is carried out in a recursive manner branch until
The corresponding data of all branch nodes correspond to identical target category.By the way that training record to be divided into purer son in succession
Collection, establishes decision tree in a recursive manner.If Dt is training record collection associated with node t, and y={ y1, y2 ..., yc } y=
{ y1, y2 ..., yc } is class label, and the recursive definition of Hunt algorithm is as follows: if all records belong to same class in Dt,
Then t is leaf node, is marked with yt.If in Dt including the record for belonging to multiple classes, an attribute test condition is selected
(attribute test condition), is divided into lesser subset for record.Each output for test condition, creation
One offspring node, and the record in Dt is distributed in children's node according to test result.Then, for every offspring section
Point recursively calls the algorithm.
Server concentrates the second feature parameter and the second target category of various kinds sheet according to test data, from test data set
In count the negative sample data of characteristic parameter combinations matches corresponding with class node each in preset rules model, counting statistics
Negative sample data shared ratio is concentrated in total negative sample data in test data, and according to calculated ratio to decision
Tree is verified.In verifying, server, which can be set, presets fault-tolerant error, when the calculated absolute difference of institute is less than default hold
It when error signal, is verified, when the calculated absolute difference of institute, which is greater than, presets fault-tolerant error, verifying does not pass through.When verifying not
By when, the sample data that server can concentrate test data is added training data and concentrates, and enlarged sample capacity is to default
Rule model is trained, and is adjusted to preset rules model.
In one of the embodiments, it is above-mentioned according to decision-tree model generate place rating model after, server according to
Place rating model regenerates passenger's rating model, so as to use passenger's rating model, but alternatively to guarantees trip
The correctness of objective rating model, server, which can preset history and reach a standard, records renewal time, and history reaches a standard record when updating
Between can be reaching a standard the time that be updated of record to security places.After arrival history, which reaches a standard, records renewal time, service
The history that device load updates reaches a standard record, and the history record that reaches a standard contains several elementary fields, including name, the age, gender with
And clearance time etc., security terminal actively or passively can send the history updated to server and reach a standard record.
Server reaches a standard from history and extracts third feature parameter and passenger's risk class in record, third feature parameter with
It is corresponding to generate the feature set in the place rating model of passenger's rating model, i.e., corresponding with target signature, passenger's risk
Grade is safety inspection result queue.
Server reaches a standard third feature parameter and passenger's risk class in record according to history, reaches a standard in record from history
The negative sample data of characteristic parameter combinations matches corresponding with class node each in each place rating model are counted, system is calculated
The negative sample data of meter reach a standard in history and record ratio shared in total negative sample data, and according to calculated ratio to each
Passenger's grade of each class node is verified in a place rating model.In verifying, server can set predetermined deviation,
When the absolute difference of the negative sample accounting of the grade in calculated ratio and passenger's grade is less than predetermined deviation, verifying is logical
It crosses；When the absolute difference of the negative sample accounting of the grade in calculated ratio and passenger's grade is greater than predetermined deviation, verifying
Do not pass through.Obstructed out-of-date when verifying, the server record that history can reach a standard continues that passenger's Rating Model is trained and is adjusted
Whole, to be reached a standard according to history, record over the ground continued to optimize by point rating model, is further carried out to passenger's rating model excellent
Change, thus by the training of big data so that the passenger's grade obtained by passenger's rating model is more and more accurate.
In above-described embodiment, place rating model is generated by decision-tree model, improves accuracy.
It should be understood that although each step in the flow chart of Fig. 2 is successively shown according to the instruction of arrow, this
A little steps are not that the inevitable sequence according to arrow instruction successively executes.Unless expressly state otherwise herein, these steps
It executes there is no the limitation of stringent sequence, these steps can execute in other order.Moreover, at least part in Fig. 2
Step may include that perhaps these sub-steps of multiple stages or stage are executed in synchronization to multiple sub-steps
It completes, but can execute at different times, the execution sequence in these sub-steps or stage, which is also not necessarily, successively to be carried out,
But it can be executed in turn or alternately at least part of the sub-step or stage of other steps or other steps.
In one embodiment, as shown in figure 3, providing a kind of passenger's rating model generating means, comprising: history reaches a standard
Record obtains module 100, grouping module 200, place rating model generation module 300 and passenger's rating model generation module 400,
Wherein:
The history record that reaches a standard obtains module 100, reaches a standard record for obtaining history, and history reaches a standard the risky trip of record carrying
Objective label or average traveler label.
Grouping module 200 is grouped for history to reach a standard to record according to place of reaching a standard.
Place rating model generation module 300, for being trained to obtain respectively pair to the record that reaches a standard of the history after grouping
The place rating model answered.
Passenger's rating model generation module 400 is commented for being combined the obtained place rating model to obtain passenger
Grade model.
Passenger's rating model generation module 400 includes: in one of the embodiments,
Receiving unit, the weight distribution for receiving input for place rating model instruct.
Weight Acquisition unit obtains the weight of place rating model for instructing according to weight distribution.
First computing unit, for according to average weighted mode base area point rating model and corresponding weight calculation
Obtain passenger's rating model.
Place rating model generation module 300 includes: in one of the embodiments,
Business rule acquiring unit for obtaining current business rule, and inquires supplement corresponding with current business rule
Characteristic parameter.
Second computing unit, for respectively according to first after grouping reach a standard record in primitive character parameter calculate supplement
Characteristic parameter.
Place rating model generation unit, for being trained to obtain pair to complementary features parameter and primitive character parameter
The place rating model answered.
Place rating model generation module 300 includes: in one of the embodiments,
Division unit, for the history record that reaches a standard after grouping to be divided into training set data and test set data.
Target's feature-extraction unit, for extracting fisrt feature parameter from training set data, according to fisrt feature parameter
Attribute gain assessment is carried out, and target signature is extracted from fisrt feature parameter according to attribute gain assessment result.
Initial rating model generation unit, for corresponding according to extracted target signature and corresponding training set data
Passenger's label training set data is classified to obtain initial rating model, and calculate each classification section in initial rating model
The grading system of point.
Feature extraction unit, for extracting second feature ginseng corresponding with selected target signature from test set data
Number.
Authentication unit, for passing through second feature parameter and the corresponding passenger's label of test set data to initial grading mould
The grading system of each node is verified to obtain the first verification result in type.
Adjustment unit obtains place rating model for being adjusted according to the first verification result to initial rating model.
Device in one of the embodiments, further include:
Loading module, for when reach history reach a standard record renewal time when, the history for loading update reaches a standard record.
Characteristic extracting module extracts corresponding with place rating model third spy in record for reaching a standard from the history of update
Levy parameter.
Authentication module records corresponding passenger's label to place for reaching a standard according to the history of third feature parameter and update
The grading system of each node is verified to obtain the second verification result in rating model.
Optimization module is optimized for putting rating model over the ground according to the second verification result.
Specific restriction about passenger's rating model generating means may refer to generate above for passenger's rating model
The restriction of method, details are not described herein.Modules in above-mentioned passenger's rating model generating means can be fully or partially through
Software, hardware and combinations thereof are realized.Above-mentioned each module can be embedded in the form of hardware or independently of the place in computer equipment
It manages in device, can also be stored in a software form in the memory in computer equipment, in order to which processor calls execution or more
The corresponding operation of modules.
In one embodiment, a kind of computer equipment is provided, which can be server, internal junction
Composition can be as shown in Figure 4.The computer equipment include by system bus connect processor, memory, network interface and
Database.Wherein, the processor of the computer equipment is for providing calculating and control ability.The memory packet of the computer equipment
Include non-volatile memory medium, built-in storage.The non-volatile memory medium is stored with operating system, computer program and data
Library.The built-in storage provides environment for the operation of operating system and computer program in non-volatile memory medium.The calculating
The database of machine equipment reaches a standard record for storing history.The network interface of the computer equipment is used to pass through with external terminal
Network connection communication.To realize a kind of passenger's rating model generation method when the computer program is executed by processor.
It will be understood by those skilled in the art that structure shown in Fig. 4, only part relevant to application scheme is tied
The block diagram of structure does not constitute the restriction for the computer equipment being applied thereon to application scheme, specific computer equipment
It may include perhaps combining certain components or with different component layouts than more or fewer components as shown in the figure.
In one embodiment, a kind of computer equipment, including memory and processor are provided, which is stored with
Computer program, which performs the steps of when executing computer program obtains history and reaches a standard record, and history reaches a standard record
Carry risk passenger label or average traveler label；History is reached a standard to record and is grouped according to place of reaching a standard；After grouping
History reach a standard record be trained to obtain corresponding place rating model respectively；Obtained place rating model is subjected to group
Conjunction obtains passenger's rating model.
In one embodiment, processor execute computer program when realized by obtained place rating model into
Row combination obtains passenger's rating model, may include: the weight distribution instruction for place rating model for receiving input；According to
Weight distribution instructs to obtain the weight of place rating model；According to average weighted mode base area point rating model and correspondence
Weight calculation obtain passenger's rating model.
In one embodiment, processor execute computer program when realized reach a standard to the history after grouping record into
Row training obtains corresponding place rating model, may include: to obtain current business rule, and inquiry is right with current business rule
The complementary features parameter answered；Respectively according to first after grouping reach a standard record in primitive character parameter calculate complementary features ginseng
Number；Complementary features parameter and primitive character parameter are trained to obtain corresponding place rating model.
In one embodiment, processor execute computer program when realized reach a standard to the history after grouping record into
Row training obtains corresponding place rating model, may include: that the record that reaches a standard of the history after grouping is divided into training set data
With test set data；Fisrt feature parameter is extracted from training set data, and attribute gain assessment is carried out according to fisrt feature parameter,
And target signature is extracted from fisrt feature parameter according to attribute gain assessment result；According to extracted target signature and right
The corresponding passenger's label of the training set data answered classifies training set data to obtain initial rating model, and calculates and initially comment
The grading system of each class node in grade model；Corresponding with selected target signature second is extracted from test set data
Characteristic parameter；By second feature parameter and the corresponding passenger's label of test set data to each node in initial rating model
Grading system verified to obtain the first verification result；Initial rating model is adjusted to obtain according to the first verification result
Place rating model.
In one embodiment, it also performs the steps of when processor executes computer program and reaches a standard note when reaching history
When recording renewal time, the history for loading update reaches a standard record；It reaches a standard from the history of update and is extracted and place rating model in recording
Corresponding third feature parameter；It is reached a standard according to the history of third feature parameter and update and records corresponding passenger's label and comment on over the ground
The grading system of each node is verified to obtain the second verification result in grade model；Grading mould is put over the ground according to the second verification result
Type optimizes.
In one embodiment, a kind of computer readable storage medium is provided, computer program is stored thereon with, is calculated
Machine program performs the steps of when being executed by processor to be obtained history and reaches a standard record, and the history record that reaches a standard carries risk passenger
Label or average traveler label；History is reached a standard to record and is grouped according to place of reaching a standard；It reaches a standard record to the history after grouping
It is trained to obtain corresponding place rating model respectively；Obtained place rating model is combined to obtain passenger's grading
Model.
In one embodiment, realized when computer program is executed by processor by obtained place rating model
It is combined to obtain passenger's rating model, may include: the weight distribution instruction for place rating model for receiving input；Root
It instructs to obtain the weight of place rating model according to weight distribution；According to average weighted mode base area point rating model and right
The weight calculation answered obtains passenger's rating model.
In one embodiment, that is realized when computer program is executed by processor reaches a standard record to the history after grouping
It is trained to obtain corresponding place rating model, may include: to obtain current business rule, and inquire and current business rule
Corresponding complementary features parameter；Respectively according to first after grouping reach a standard record in primitive character parameter calculate complementary features ginseng
Number；Complementary features parameter and primitive character parameter are trained to obtain corresponding place rating model.
In one embodiment, that is realized when computer program is executed by processor reaches a standard record to the history after grouping
It is trained to obtain corresponding place rating model, may include: that the record that reaches a standard of the history after grouping is divided into training set number
According to test set data；Fisrt feature parameter is extracted from training set data, and attribute gain is carried out according to fisrt feature parameter and is commented
Estimate, and target signature is extracted from fisrt feature parameter according to attribute gain assessment result；According to extracted target signature with
And the corresponding passenger's label of corresponding training set data classifies training set data to obtain initial rating model, and calculates just
The grading system of each class node in beginning rating model；It is extracted from test set data corresponding with selected target signature
Second feature parameter；By second feature parameter and the corresponding passenger's label of test set data to each in initial rating model
The grading system of node is verified to obtain the first verification result；Initial rating model is adjusted according to the first verification result
Obtain place rating model.
In one embodiment, it is also performed the steps of when computer program is executed by processor when arrival history reaches a standard
When recording renewal time, the history for loading update reaches a standard record；It reaches a standard and is extracted in recording and grading mould in place from the history of update
The corresponding third feature parameter of type；It is reached a standard according to the history of third feature parameter and update and records corresponding passenger's label to place
The grading system of each node is verified to obtain the second verification result in rating model；Put grading over the ground according to the second verification result
Model optimizes.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the computer program can be stored in a non-volatile computer
In read/write memory medium, the computer program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein,
To any reference of memory, storage, database or other media used in each embodiment provided herein,
Including non-volatile and/or volatile memory.Nonvolatile memory may include read-only memory (ROM), programming ROM
(PROM), electrically programmable ROM (EPROM), electrically erasable ROM (EEPROM) or flash memory.Volatile memory may include
Random access memory (RAM) or external cache.By way of illustration and not limitation, RAM is available in many forms,
Such as static state RAM (SRAM), dynamic ram (DRAM), synchronous dram (SDRAM), double data rate sdram (DDRSDRAM), enhancing
Type SDRAM (ESDRAM), synchronization link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM
(RDRAM), direct memory bus dynamic ram (DRDRAM) and memory bus dynamic ram (RDRAM) etc..
Each technical characteristic of above embodiments can be combined arbitrarily, for simplicity of description, not to above-described embodiment
In each technical characteristic it is all possible combination be all described, as long as however, the combination of these technical characteristics be not present lance
Shield all should be considered as described in this specification.
The several embodiments of the application above described embodiment only expresses, the description thereof is more specific and detailed, but simultaneously
It cannot therefore be construed as limiting the scope of the patent.It should be pointed out that coming for those of ordinary skill in the art
It says, without departing from the concept of this application, various modifications and improvements can be made, these belong to the protection of the application
Range.Therefore, the scope of protection shall be subject to the appended claims for the application patent.
Claims (10)
1. a kind of passenger's rating model generation method, which comprises
It obtains history to reach a standard record, the history record that reaches a standard carries risk passenger label or average traveler label；
The history is reached a standard to record and is grouped according to place of reaching a standard；
It reaches a standard to record to the history after grouping and is trained to obtain corresponding place rating model respectively；
It is combined obtained place rating model to obtain passenger's rating model.
2. the method according to claim 1, wherein described be combined obtained place rating model
To passenger's rating model, comprising:
The weight distribution for the place rating model for receiving input instructs；
It is instructed to obtain the weight of the place rating model according to the weight distribution；
Passenger's rating model is obtained according to the place rating model and corresponding weight calculation according to average weighted mode.
3. record is trained the method according to claim 1, wherein the history after described pair of grouping reaches a standard
To corresponding place rating model, comprising:
Current business rule is obtained, and inquires complementary features parameter corresponding with the current business rule；
Respectively according to described first after grouping reach a standard record in primitive character parameter calculate complementary features parameter；
The complementary features parameter and primitive character parameter are trained to obtain corresponding place rating model.
4. according to claim 1 to method described in 3 any one, which is characterized in that the history after described pair of grouping reaches a standard note
Record is trained to obtain corresponding place rating model, comprising:
The history after the grouping record that reaches a standard is divided into training set data and test set data；
Fisrt feature parameter is extracted from the training set data, and attribute gain assessment is carried out according to the fisrt feature parameter,
And target signature is extracted from the fisrt feature parameter according to attribute gain assessment result；
According to extracted target signature and the corresponding passenger's label of the corresponding training set data to the training set number
According to being classified to obtain initial rating model, and calculate the grading system of each class node in the initial rating model；
Second feature parameter corresponding with selected target signature is extracted from the test set data；
By the second feature parameter and the corresponding passenger's label of the test set data in the initial rating model
The grading system of each node is verified to obtain the first verification result；
The initial rating model is adjusted to obtain place rating model according to first verification result.
5. according to the method described in claim 4, it is characterized in that, the method also includes:
When reach history reach a standard record renewal time when, the history for loading update reaches a standard record；
It reaches a standard from the history of update and extracts third feature parameter corresponding with the place rating model in recording；
It is reached a standard according to the history of the third feature parameter and update and records corresponding passenger's label and grade to the place
The grading system of each node is verified to obtain the second verification result in model；
The place rating model is optimized according to second verification result.
6. a kind of passenger's rating model generating means, which is characterized in that described device includes:
The history record that reaches a standard obtains module, reaches a standard record for obtaining history, and the history record that reaches a standard carries risk passenger
Label or average traveler label；
Grouping module is grouped for the history to reach a standard to record according to place of reaching a standard；
Place rating model generation module is trained to obtain corresponding place respectively for reaching a standard to record to the history after grouping
Rating model；
Passenger's rating model generation module, for being combined obtained place rating model to obtain passenger's rating model.
7. device according to claim 6, which is characterized in that passenger's rating model generation module includes:
Receiving unit, the weight distribution for receiving input for the place rating model instruct；
Weight Acquisition unit obtains the weight of the place rating model for instructing according to the weight distribution；
First computing unit is used for according to average weighted mode according to the place rating model and corresponding weight calculation
Obtain passenger's rating model.
8. device according to claim 6, which is characterized in that the place rating model generation module includes:
Business rule acquiring unit for obtaining current business rule, and inquires supplement corresponding with the current business rule
Characteristic parameter；
Second computing unit, for respectively according to described first after grouping reach a standard record in primitive character parameter calculate supplement
Characteristic parameter；
Place rating model generation unit, for being trained to obtain pair to the complementary features parameter and primitive character parameter
The place rating model answered.
9. a kind of computer equipment, including memory and processor, the memory are stored with computer program, feature exists
In the step of processor realizes any one of claims 1 to 5 the method when executing the computer program.
10. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program
The step of method described in any one of claims 1 to 5 is realized when being executed by processor.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810788417.7A CN109102159B (en) | 2018-07-18 | 2018-07-18 | Passenger rating model generation method, device, computer equipment and storage medium |
| PCT/CN2018/106083 WO2020015140A1 (en) | 2018-07-18 | 2018-09-18 | Passenger rating model generation method and apparatus, and computer device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810788417.7A CN109102159B (en) | 2018-07-18 | 2018-07-18 | Passenger rating model generation method, device, computer equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109102159A true CN109102159A (en) | 2018-12-28 |
| CN109102159B CN109102159B (en) | 2023-06-20 |
Family
ID=64846639
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810788417.7A Active CN109102159B (en) | 2018-07-18 | 2018-07-18 | Passenger rating model generation method, device, computer equipment and storage medium |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN109102159B (en) |
| WO (1) | WO2020015140A1 (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110598995A (en) * | 2019-08-15 | 2019-12-20 | 中国平安人寿保险股份有限公司 | Intelligent customer rating method and device and computer readable storage medium |
| CN111352171A (en) * | 2020-03-30 | 2020-06-30 | 重庆特斯联智慧科技股份有限公司 | Method and system for realizing artificial intelligence regional shielding security inspection |
| CN113052689A (en) * | 2021-04-30 | 2021-06-29 | 中国银行股份有限公司 | Product recommendation method and device based on decision tree |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111831904A (en) * | 2020-06-18 | 2020-10-27 | 天讯瑞达通信技术有限公司 | Passenger behavior data analysis method and system |
| CN115001771B (en) * | 2022-05-25 | 2024-01-26 | 武汉极意网络科技有限公司 | Verification code defending method, system, equipment and storage medium based on automatic updating |
| CN118337532B (en) * | 2024-06-13 | 2024-08-23 | 浙江鹏信信息科技股份有限公司 | Zero trust-based traffic safety audit protection method and system and readable medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106030626A (en) * | 2013-12-11 | 2016-10-12 | 天巡有限公司 | Method and system for providing fare availabilities, such as air fare availabilities |
| CN106874951A (en) * | 2017-02-14 | 2017-06-20 | Tcl集团股份有限公司 | A kind of passenger's attention rate ranking method and device |
| CN107194412A (en) * | 2017-04-20 | 2017-09-22 | 百度在线网络技术（北京）有限公司 | A kind of method of processing data, device, equipment and computer-readable storage medium |
| CN107590569A (en) * | 2017-09-25 | 2018-01-16 | 山东浪潮云服务信息科技有限公司 | A kind of data predication method and device |
| WO2018090839A1 (en) * | 2016-11-16 | 2018-05-24 | 阿里巴巴集团控股有限公司 | Identity verification system, method, device, and account verification method |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106570631B (en) * | 2016-10-28 | 2021-01-01 | 南京邮电大学 | P2P platform-oriented operation risk assessment method and system |
| CN108269012A (en) * | 2018-01-12 | 2018-07-10 | 中国平安人寿保险股份有限公司 | Construction method, device, storage medium and the terminal of risk score model |
-
2018
- 2018-07-18 CN CN201810788417.7A patent/CN109102159B/en active Active
- 2018-09-18 WO PCT/CN2018/106083 patent/WO2020015140A1/en active Application Filing
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106030626A (en) * | 2013-12-11 | 2016-10-12 | 天巡有限公司 | Method and system for providing fare availabilities, such as air fare availabilities |
| WO2018090839A1 (en) * | 2016-11-16 | 2018-05-24 | 阿里巴巴集团控股有限公司 | Identity verification system, method, device, and account verification method |
| CN106874951A (en) * | 2017-02-14 | 2017-06-20 | Tcl集团股份有限公司 | A kind of passenger's attention rate ranking method and device |
| CN107194412A (en) * | 2017-04-20 | 2017-09-22 | 百度在线网络技术（北京）有限公司 | A kind of method of processing data, device, equipment and computer-readable storage medium |
| CN107590569A (en) * | 2017-09-25 | 2018-01-16 | 山东浪潮云服务信息科技有限公司 | A kind of data predication method and device |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110598995A (en) * | 2019-08-15 | 2019-12-20 | 中国平安人寿保险股份有限公司 | Intelligent customer rating method and device and computer readable storage medium |
| CN110598995B (en) * | 2019-08-15 | 2023-08-25 | 中国平安人寿保险股份有限公司 | Smart client rating method, smart client rating device and computer readable storage medium |
| CN111352171A (en) * | 2020-03-30 | 2020-06-30 | 重庆特斯联智慧科技股份有限公司 | Method and system for realizing artificial intelligence regional shielding security inspection |
| CN113052689A (en) * | 2021-04-30 | 2021-06-29 | 中国银行股份有限公司 | Product recommendation method and device based on decision tree |
| CN113052689B (en) * | 2021-04-30 | 2024-03-26 | 中国银行股份有限公司 | Product recommendation method and device based on decision tree |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2020015140A1 (en) | 2020-01-23 |
| CN109102159B (en) | 2023-06-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109102159A (en) | Passenger's rating model generation method, device, computer equipment and storage medium | |
| CN108564129B (en) | Trajectory data classification method based on generation countermeasure network | |
| CN109242740A (en) | Identity information risk assessment method, apparatus, computer equipment and storage medium | |
| CN112989035B (en) | Method, device and storage medium for identifying user intention based on text classification | |
| CN109063984B (en) | Method, apparatus, computer device and storage medium for risky travelers | |
| CN106503863A (en) | Based on the Forecasting Methodology of the age characteristicss of decision-tree model, system and terminal | |
| CN111797320B (en) | Data processing method, device, equipment and storage medium | |
| CN109886554A (en) | Unlawful practice method of discrimination, device, computer equipment and storage medium | |
| CN109325542A (en) | A kind of electricity exception intelligent identification Method and system based on multistage machine learning | |
| CN116415206B (en) | Operator multiple data fusion method, system, electronic equipment and computer storage medium | |
| CN112365007B (en) | Model parameter determining method, device, equipment and storage medium | |
| CN110069545A (en) | A kind of behavioral data appraisal procedure and device | |
| CN111767962A (en) | One-stage target detection method, system and device based on generation countermeasure network | |
| CN108198172A (en) | Image significance detection method and device | |
| CN110929806A (en) | Picture processing method and device based on artificial intelligence and electronic equipment | |
| CN113822315A (en) | Attribute graph processing method and device, electronic equipment and readable storage medium | |
| CN112396428B (en) | User portrait data-based customer group classification management method and device | |
| CN111126264A (en) | Image processing method, device, equipment and storage medium | |
| CN111210158B (en) | Target address determining method, device, computer equipment and storage medium | |
| CN114330650A (en) | Small sample characteristic analysis method and device based on evolutionary element learning model training | |
| CN111275059B (en) | Image processing method and device and computer readable storage medium | |
| CN111325255B (en) | Specific crowd delineating method and device, electronic equipment and storage medium | |
| CN113065662A (en) | Data processing method, self-learning system and electronic equipment | |
| CN110210425A (en) | Face identification method, device, electronic equipment and storage medium | |
| CN116910341A (en) | Label prediction method and device and electronic equipment |
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 |