CN109360105A - Product risks method for early warning, device, computer equipment and storage medium - Google Patents
Product risks method for early warning, device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN109360105A CN109360105A CN201811089567.5A CN201811089567A CN109360105A CN 109360105 A CN109360105 A CN 109360105A CN 201811089567 A CN201811089567 A CN 201811089567A CN 109360105 A CN109360105 A CN 109360105A
- Authority
- CN
- China
- Prior art keywords
- product
- data
- analyzed
- sample
- risk
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 41
- 239000013598 vector Substances 0.000 claims abstract description 160
- 238000003062 neural network model Methods 0.000 claims abstract description 80
- 238000012545 processing Methods 0.000 claims abstract description 44
- 238000004590 computer program Methods 0.000 claims description 29
- 238000011156 evaluation Methods 0.000 claims description 18
- 238000013139 quantization Methods 0.000 claims description 8
- 238000007405 data analysis Methods 0.000 claims description 4
- 238000004519 manufacturing process Methods 0.000 claims description 3
- 210000005036 nerve Anatomy 0.000 claims 1
- 238000012502 risk assessment Methods 0.000 abstract description 11
- 241001269238 Data Species 0.000 abstract description 3
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 239000000047 product Substances 0.000 description 233
- 238000012549 training Methods 0.000 description 17
- 238000004458 analytical method Methods 0.000 description 11
- 238000013528 artificial neural network Methods 0.000 description 11
- 238000010586 diagram Methods 0.000 description 5
- 210000004218 nerve net Anatomy 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000010276 construction Methods 0.000 description 2
- 235000013399 edible fruits Nutrition 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 230000015572 biosynthetic process Effects 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 239000006227 byproduct Substances 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000010224 classification analysis Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011002 quantification Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012795 verification Methods 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
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/06—Asset management; Financial planning or analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Finance (AREA)
- Development Economics (AREA)
- Accounting & Taxation (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Entrepreneurship & Innovation (AREA)
- Health & Medical Sciences (AREA)
- Game Theory and Decision Science (AREA)
- Human Resources & Organizations (AREA)
- Operations Research (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Technology Law (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
This application involves artificial intelligence fields, can be applied to financial industry, provide a kind of product risks method for early warning, device, computer equipment and storage medium.Method includes: to obtain product information to be analyzed, the processing of data portrait is carried out to product information to be analyzed, obtain the corresponding vector data of product information to be analyzed, vector data is combined with multiple sample vector data of preset twin neural network model respectively, obtained multipair data splitting is inputted into preset twin neural network model, the risk probability for obtaining product to be analyzed pushes the Risk-warning information of product to be analyzed according to risk probability.This method is handled by data portrait, the related data of product to be analyzed can deeply be excavated, it is analysed to the vector data of product and sample vector data is combined as input data, and the similarity of two input datas is evaluated using preset twin neural network model, keep the risk analysis result obtained and corresponding warning information more accurate effectively.
Description
Technical field
This application involves field of artificial intelligence, more particularly to a kind of product risks method for early warning, device, computer
Equipment and storage medium.
Background technique
With the development of financial industry, occur more and more being similar to the financial investment class product of bond, such as entreat
Enterprise, state-owned enterprise, Deng Ge major class enterprise of private enterprise such as can also issue bond at the products, and investor is when selecting investment product, for protection
Number one avoids property loss, generally can all tend to the lesser investment product of investment risk.
It is traditional to need to use a large amount of historical data for the risk analysis method of investment product and test and assess, determine it
Potential risks, but in a practical situation, the available negative sample accounting for violation of agreement occur and its small in bond data,
Traditional method is difficult to realize effective risk analysis to product.
Summary of the invention
Based on this, it is necessary in view of the above technical problems, provide a kind of product wind of raising product risks analysis validity
Dangerous method for early warning, device, computer equipment and storage medium.
A kind of product risks method for early warning, which comprises
Product information to be analyzed is obtained, the processing of data portrait is carried out to the product information to be analyzed, is obtained described wait divide
Analyse the corresponding vector data of product information；
The sample vector data for obtaining preset twin neural network model, by the vector data respectively with it is multiple described
Sample vector data are combined, and obtain multipair data splitting；
The multipair data splitting is inputted into the preset twin neural network model, obtains the risk of product to be analyzed
Probability；
According to the risk probability, the Risk-warning information of the product to be analyzed is pushed.
It is described in one of the embodiments, to obtain product information to be analyzed, the product information to be analyzed is counted
It is handled according to portrait, obtains the corresponding vector data of the product information to be analyzed, comprising:
Each dimensional information of product to be analyzed is obtained, and according to default quantitative criteria, obtains the amount of each dimensional information
Change data；
The processing of data portrait is carried out to the quantized data of each dimensional information, obtains the generaton number of the product to be analyzed
According to；
According to the quantized data and the derivative data, the vector data of the product to be analyzed is obtained.
Product risks method for early warning in one of the embodiments, further include:
The multipair data splitting is inputted into the preset twin neural network model, obtains the highest combination of similarity
Data；
According to the highest data splitting of the similarity, sample vector data corresponding with the product to be analyzed are obtained；
According to the risk situation of the sample vector data corresponding product, the potential risk feelings of the product to be analyzed are obtained
Condition.
It is described in one of the embodiments, that the multipair data splitting is inputted into the preset twin neural network mould
Type, before the risk probability for obtaining product to be analyzed, further includes:
Sample product is obtained, the sample product is subjected to the processing of data portrait, obtains sample vector data；
The sample product is analyzed, determines the corresponding data label of the sample vector data；
According to the data label, processing is grouped in the sample vector data, acquisition group is to sample；
According to described group to sample, building obtains preset twin neural network model.
In one of the embodiments, described group to sample include positive group to sample with negative group to sample；It is described according to institute
Data label is stated, processing is grouped in the sample vector data, acquisition group is to sample, comprising:
According to the data label that sample vector data carry, sample vector data are classified；
According to permutation and combination, same category of two sample vector data are grouped, obtain positive group to sample；
According to permutation and combination, two different classes of sample vector data are grouped, obtain negative group to sample.
In one of the embodiments, it is described according to described group to sample, building obtains preset twin neural network mould
Type includes:
According to described group to sample, initial twin neural network model is generated；
Obtain the model evaluation parameter of the initial twin neural network model；
When the model evaluation parameter is not up to preset threshold range, by back-propagation algorithm to described initial twin
The structural parameters of neural network model are adjusted, and building obtains the preset twin neural network model.
It is described according to the risk probability in one of the embodiments, push the Risk-warning of the product to be analyzed
Information, comprising:
According to product alert demand to be analyzed, Risk-warning grade threshold is determined；
According to the risk probability and the Risk-warning grade threshold, the corresponding risk of the product to be analyzed is determined
Warning grade, and push the Risk-warning information of the product to be analyzed.
A kind of product risks prior-warning device, described device include:
Data processing module carries out data portrait to the product information to be analyzed for obtaining product information to be analyzed
Processing obtains the corresponding vector data of the product information to be analyzed；
Data combination module, for obtaining the sample vector data of preset twin neural network model, by the vector
Data are combined with multiple sample vector data respectively, obtain multipair data splitting；
Data analysis module is obtained for the multipair data splitting to be inputted the preset twin neural network model
Obtain the risk probability of product to be analyzed；
Risk-warning module, for pushing the Risk-warning information of the product to be analyzed according to the risk probability.
A kind of computer equipment, including memory and processor, the memory are stored with computer program, the processing
Device performs the steps of when executing the computer program
Product information to be analyzed is obtained, the processing of data portrait is carried out to the product information to be analyzed, is obtained described wait divide
Analyse the corresponding vector data of product information；
The sample vector data for obtaining preset twin neural network model, by the vector data respectively with it is multiple described
Sample vector data are combined, and obtain multipair data splitting；
The multipair data splitting is inputted into the preset twin neural network model, obtains the risk of product to be analyzed
Probability；
According to the risk probability, the Risk-warning information of the product to be analyzed is pushed.
A kind of computer readable storage medium, is stored thereon with computer program, and the computer program is held by processor
It is performed the steps of when row
Product information to be analyzed is obtained, the processing of data portrait is carried out to the product information to be analyzed, is obtained described wait divide
Analyse the corresponding vector data of product information；
The sample vector data for obtaining preset twin neural network model, by the vector data respectively with it is multiple described
Sample vector data are combined, and obtain multipair data splitting；
The multipair data splitting is inputted into the preset twin neural network model, obtains the risk of product to be analyzed
Probability；
According to the risk probability, the Risk-warning information of the product to be analyzed is pushed.
The said goods method for prewarning risk, device, computer equipment and storage medium are believed by obtaining product to be analyzed
Breath is handled product information to be analyzed using data portrait, obtains the vector data for characterizing product to be analyzed, and will
Vector data is combined with sample vector data, and combined result is inputted preset twin neural network model, is obtained wait divide
The risk probability of division product can be with so that it is determined that the Risk-warning grade of product to be analyzed, the application are handled by data portrait
The related data for deeply excavating product to be analyzed, is analysed to the vector data of product and sample data is combined as input number
According to, and evaluate using preset twin neural network model the similarities of two inputs, make the risk analysis result obtained with
And corresponding warning information is more accurate effectively.
Detailed description of the invention
Fig. 1 is the flow diagram of product risks method for early warning in one embodiment；
Fig. 2 is the flow diagram of product risks method for early warning in another embodiment；
Fig. 3 is the flow diagram of product risks method for early warning in another embodiment；
Fig. 4 is the structural block diagram of product risks prior-warning device in one embodiment；
Fig. 5 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.
Product risks method for early warning provided by the present application is analyzed and is pushed for the risk situation to financial products
Warning information can specifically be realized, computer program by product risks method for early warning of the computer program to the application
It can load in terminal, terminal can be, but not limited to be various personal computers, laptop, smart phone, plate electricity
Brain.
In one embodiment, as shown in Figure 1, providing a kind of product risks method for early warning, comprising the following steps:
Step S200 obtains product information to be analyzed, carries out the processing of data portrait to product information to be analyzed, obtains wait divide
Analyse the corresponding vector data of product information.
Product to be analyzed refers to that there may be financial class investment product, such as bond, fund of risk etc., productions to be analyzed
Product information refers to each dimensional information relevant to product to be analyzed, for example, the bond face amount of bond product, maturity, interest payment period,
The information such as coupon rate, the processing of data portrait refer to that being analysed to product information carries out the processing of various dimensions portrait, obtains to be analyzed
The process of the profound information of product can convert the text information of acquisition to the data of quantization in portrait treatment process
Information is analyzed by data, obtains the vector data of product to be analyzed, can more comprehensive intuitively body by vector data
The feature of existing product to be analyzed.
Step S300 obtains the sample vector data of preset twin neural network model, by vector data respectively with it is more
A sample vector data are combined, and obtain multipair data splitting.
Sample vector data refer to the vector number by the product of known risk situation, obtained after the processing of data portrait
It is the training data of preset twin neural network model according to, sample vector data there are multiple, it can be by the contingency table of setting
Standard is stored in corresponding sample vector database, carries out group by the vector data and sample vector data that are analysed to product
It closes, multiple groups can be obtained for the data splitting of similarity-rough set as input data, to pass through preset twin nerve net
The determination of network model and the highest sample vector data of product similarity to be analyzed, and obtain the risk probability of product to be analyzed.
Multipair data splitting is inputted preset twin neural network model, obtains the wind of product to be analyzed by step S400
Dangerous probability.
Twin neural network refers to that a kind of includes the neural network framework of two or more identical sub-networks, here phase
With referring to their configuration having the same parameters i.e. having the same and weight.Parameter update carries out jointly on the net in two sons,
Weight is shared by subnet, so that the training of model needs less sample data, also means that and needs less data simultaneously
And it is not easy over-fitting, twin neural network is being related to finding the task of the relationship between similitude or two comparable things
In get application.Such as scoring task is repeated, wherein input is two sentences, output is the similarity degree of two sentences, is determined
The negative score repeated.Such as signature verification task, it is analyzed by twin neural network and determines two signatures whether from same
It is personal.In general, handling two inputs using two identical sub-networks, and according to two sub-networks in such task
Output generate final output.Using twin neural network model, the similarity degree for obtaining each pair of data splitting can analyze, and
It is analyzed to obtain the corresponding risk probability of each data according to established standards.
Step S500 pushes the Risk-warning information of product to be analyzed according to risk probability.
Risk probability refers to the probability according to the product promise breaking exported after preset twin neural network model analysis processing
Size relatively determines that the product whether there is a degree of risk, and tie according to comparing by the way that Risk-warning threshold value is arranged
Fruit pushes the Risk-warning information of the product.Risk-warning information may include specific risk analysis as a result, as the product which
Risk existing for aspect is larger, and the potential violation of agreement of this aspect is as how.
The said goods method for prewarning risk is drawn a portrait using data to product to be analyzed by obtaining product information to be analyzed
Information is handled, and obtains the vector data for characterizing product to be analyzed, and vector data and sample vector data are carried out
Combination, inputs preset twin neural network model for combined result, obtains the risk probability of product to be analyzed, so that it is determined that
The Risk-warning grade of product is analyzed, the application is handled by data portrait, can deeply excavate the dependency number of product to be analyzed
According to being analysed to the vector data of product and sample data be combined as input data, and utilize preset twin nerve net
Network model evaluates the similarities of two inputs, makes the risk analysis result obtained and corresponding warning information is more accurate has
Effect.
In one embodiment, as shown in Fig. 2, step S200, obtains product information to be analyzed, to product information to be analyzed
The processing of data portrait is carried out, the corresponding vector data of product information to be analyzed is obtained, comprising:
Step S220 obtains each dimensional information of product to be analyzed, and according to default quantitative criteria, obtains each dimensional information
Quantized data.
Step S240 carries out the processing of data portrait to the quantized data of each dimensional information, obtains the derivative of product to be analyzed
Data.
Step S260 obtains the vector data of product to be analyzed according to quantized data and derivative data.
Each investment product to be analyzed can have the product related news of various dimensions, for example, bond product information
Dimension includes but is not limited to bond face amount, maturity, interest payment period, coupon rate etc., and preset quantitative criteria refers to by letter
After the parsing for ceasing content, the criterion of quantized data corresponding to the information passes through by taking the par value information of credits product as an example
Bond face amount is compared with given threshold, be information quantization within the set range by bond face amount is 1, not in range
Information quantization be 0 or other quantized datas, and according to identical rule, quantification treatment is carried out to the information of each dimension.It will
Data conversion after each dimension quantization is indicated at matrix, is deeply divided matrix according to different dimensions using Portrait brand technology
Analysis, obtains the derivative data of product to be analyzed, and quantized data and derivative data are arranged according to certain requirement, formed bond to
Amount.
In one embodiment, as shown in figure 3, product risks method for early warning, further includes:
Multipair data splitting is input to preset twin neural network model, it is highest to obtain similarity by step S420
Data splitting.
Step S460 obtains sample vector data corresponding with product to be analyzed according to the highest data splitting of similarity.
Step S480 obtains the potential risk of product to be analyzed according to the risk situation of sample vector data corresponding product.
The basic ideas of twin neural network model are that analysis obtains the similitude of two input datas, by multipair number of combinations
According to twin neural network model is input to, twin neural network model can be analyzed to obtain a group analysis for every a pair of of data splitting
As a result, analyzing the comparison of result by multiple groups, the risk probability of product to be analyzed and the highest data splitting of similarity are determined
Corresponding result.It include the vector data and a sample vector data of sample to be analysed in data splitting, according to twin mind
Analysis through pessimistic concurrency control is as a result, sample vector number corresponding with product to be analyzed in the highest data splitting of similarity can be determined
Be the corresponding vector data of sample product of known risk situation according to, sample vector data, due to sample product similarity with
The similarity of product to be analyzed is very high, by the risk situation of the sample product, is mapped by comparison, can obtain production to be analyzed
The potential risk of product, potential risk refer to the default risk situation that product is likely to occur, for example, with product similarity to be analyzed
Within listing half a year violation of agreement occurs for high sample product, is that there may be disobey in half a year in the potential risk of product to be analyzed
About.According to risk probability and potential risk, the Risk-warning information of product to be analyzed is generated, and pushes to user.
In one embodiment, multipair data splitting is inputted preset twin neural network model, obtained by step S400
Before the risk probability of product to be analyzed, further includes:
Sample product is obtained, sample product is subjected to the processing of data portrait, obtains sample vector data.
Sample product is analyzed, determines the corresponding data label of sample vector data.
According to data label, processing is grouped in sample vector data, acquisition group is to sample.
According to group to sample, building obtains preset twin neural network model.
Sample product can be the known product for having occurred and that the known product of promise breaking or not breaking a contract, by sample
The data portrait of the various dimensions information of product is handled, and the quantized data and derivative data of the sample product is obtained, to obtain sample
The vector data of this product.Since sample product is known product, classification analysis is carried out by the Given information to known product,
It can determine that the corresponding data label of the sample product is taken by the vector data of the sample product in conjunction with data label
Vector data with data label, data label are the characteristic parameters for characterizing vector data.Data label may include multilayer
Subtab, wherein the subtab of sample product can be determined by product classification, for example, bond information can according to market
To divide issuing market and circulation market, bond circulation market can be further divided into transaction on exchange market and over-the-counter trading city again
?；Bond is divided into default bond or non-default bond according to the situation of honouring an agreement of bond, and further will according to promise breaking mode
Default bond is divided into long-term bond promise breaking and short term bond promise breaking.
In one of the embodiments, group to sample include positive group to sample with negative group to sample.It, will according to data label
Processing is grouped in sample vector data, and acquisition group is to sample, comprising:
According to the data label that sample vector data carry, sample vector data are classified.
According to permutation and combination, same category of two sample vector data are grouped, obtain positive group to sample.
According to permutation and combination, two different classes of sample vector data are grouped, obtain negative group to sample.
The training process of network model often relies on training sample, and in fact, most of the available of data analysis is born
Sample is few, and negative sample here refers to the sample that violation of agreement occurs in such as bond product.Use twin neural network mould
Type, using its training sample group to input the characteristics of, sample combine by way of, expand sample size, solve available
The few problem of sample.Sample vector data carry data label, sample vector data can be classified, wherein can root
Sample bond product is divided into the sample and not of having broken a contract according to the categorical measure for needing to set classification, such as according to whether promise breaking occurs
Promise breaking two class of sample.It is appreciated that in other embodiments, can also determine class categories according to the quantity of sample vector data
Quantity, when total sample is less, it is possible to reduce class categories can increase class categories when total sample is more.Group pair
Sample include positive group to sample with negative group to sample, wherein positive group refers to the group of same category of two samples to knot in sample
Fruit, negative group refers to the group pair of two different classes of samples to sample as a result, in embodiment, can pass through the side of permutation and combination
The group pair of formula realization sample.Sample is trained sample to model with negative group using positive group, compares and only with positive group to sample
For this progress model training, preferable training result can be obtained, help to obtain the preferably twin neural network mould of effect
Type.
In one of the embodiments, according to group to sample, building obtains preset twin neural network model and includes:
According to group to sample, initial twin neural network model is generated；
Obtain the model evaluation parameter of initial twin neural network model；
When model evaluation parameter is not up to preset threshold range, by back-propagation algorithm to initial twin neural network
The structural parameters of model are adjusted, and building obtains preset twin neural network model.
After the completion of model training, it is also necessary to evaluate the twin neural network model after training, judge that it is accurate
Whether the evaluation parameters such as rate reach expected, i.e., the output result that the twin neural network model that training of judgement is completed is analyzed
Whether preset acceptable range judges that its training is completed, can be used for treating when it is in preset error range
Analysis product is analyzed.In embodiment, evaluation parameter includes accuracy rate, accuracy, recall rate and F (F-Measure)
The model evaluations parameter such as value is adjusted, wherein accuracy rate refers to the sample number that meets the requirements of output result divided by all samples
This number, usually, accuracy are higher, and modelling effect is better.It is actually positive example that accuracy, which refers in the example for be divided into positive example,
Ratio, can be used to measured model to the recognition capability of positive example.Recall rate is the measurement of covering surface, and measurement has multiple
Positive example is divided into positive example.F value refers to comprehensive evaluation index, is the weighted harmonic mean to accuracy and recall rate, when F value is higher
Shi Zeneng illustrates that model is more effective.When model evaluation parameter is not up to claimed range, to twin by way of backpropagation
The parameter of raw neural network model is adjusted.
In one embodiment, step S500 pushes the Risk-warning information of product to be analyzed according to risk probability, packet
It includes:
According to product alert demand to be analyzed, Risk-warning grade threshold is determined.
According to risk probability and Risk-warning grade threshold, the corresponding Risk-warning grade of product to be analyzed is determined, and
Push the Risk-warning information of product to be analyzed.
Product alert demand to be analyzed can be used for characterizing the unacceptable risk range of product to be analyzed and unacceptable
Degree, due to different product for risk sensitivity there may be difference, can be with according to the early warning demand of product to be analyzed
Dangerous warning grade threshold value is reasonably set, divides different Risk-warning grades, is such as divided into high risk, risk, low-risk
Three grades, wherein the number of grade can be carried out threshold value setting according to different needs by user to determine.When by twin
When the risk probability that neural network model obtains is more than minimum risk threshold value, wind corresponding to the risk probability is further judged
Dangerous warning grade, and it is analysed to the warning grade where product and relevant potential risk situation pushes to user.
In an application example, the product risks method for early warning of the application be used to carry out bond product risk analysis with
Early warning processing, comprising: by obtaining multiple known bond product informations, using Portrait brand technology from multiple dimensions to bond information
The processing of data portrait is carried out, quantization generates bond vector；And bond is analyzed, determine the corresponding data mark of bond vector
Bond vector is divided into positive sample and negative sample according to data label by rifle, and sample is carried out two by the way of permutation and combination
Two combinations, realize the expansion of sample size, obtain the combination of sample vector data；Input sample vector data combined training carries out mould
The training of twin neural network model is realized in type training, after the completion of the training of twin neural network model, is joined by model evaluation
Whether the model after number assessment training is meeting the requirements, when being unsatisfactory for demand, by back-propagation algorithm to parameter under model
It optimizes, when meet demand, twin neural network model can be used for analyzing the risk situation of new bond, first to new
Bond carry out the processing of data portrait, obtain new bond vector, and by new bond vector and existing sample vector data
Vector is combined, and the multipair combination of formation is inputted twin neural network model, according to the calculating of twin neural network model
Default Probability is obtained, according to the threshold range where Default Probability, determines the risk class of the bond, and according to twin nerve net
The combination similarity of network model output, determining and highest existing risk bonds of new bond similarity, thus according to this
The risk situation of risky bond determines the potential risk of new bond product.
It should be understood that although each step in the flow chart of Fig. 1-3 is successively shown according to the instruction of arrow,
These steps are not that the inevitable sequence according to arrow instruction successively executes.Unless expressly stating otherwise herein, these steps
Execution there is no stringent sequences to limit, these steps can execute in other order.Moreover, at least one in Fig. 1-3
Part steps may include that perhaps these sub-steps of multiple stages or stage are not necessarily in synchronization to multiple sub-steps
Completion is executed, but can be executed at different times, the execution sequence in these sub-steps or stage is also not necessarily successively
It carries out, but can be at least part of the sub-step or stage of other steps or other steps in turn or alternately
It executes.
In one embodiment, as shown in figure 4, providing a kind of product risks prior-warning device, comprising:
Data processing module 200 carries out at data portrait product information to be analyzed for obtaining product information to be analyzed
Reason, obtains the corresponding vector data of product information to be analyzed；
Data combination module 300, for obtaining the sample vector data of preset twin neural network model, by vector number
According to being combined respectively with multiple sample vector data, multipair data splitting is obtained；
Data analysis module 400 is obtained for multipair data splitting to be inputted preset twin neural network model wait divide
The risk probability of division product；
Early warning pushing module 500, for pushing the Risk-warning information of product to be analyzed according to risk probability.
In one embodiment, data processing module 200, are also used to obtain each dimensional information of product to be analyzed, and press
According to default quantitative criteria, the quantized data of each dimensional information is obtained, the quantized data of each dimensional information is carried out at data portrait
Reason, obtains the derivative data of product to be analyzed, according to quantized data and derivative data, obtains the vector data of product to be analyzed.
In one embodiment, product risks prior-warning device further includes potential risk analysis module, is used for multipair combination
Data are input to preset twin neural network model, the highest data splitting of similarity are obtained, according to highest group of similarity
Data are closed, corresponding with product to be analyzed sample vector data are obtained, according to the risk situation of sample vector data corresponding product,
Obtain the potential risk of product to be analyzed.
In one embodiment, product risks prior-warning device further includes model construction module, will for obtaining sample product
Sample product carries out the processing of data portrait, obtains sample vector data, analyzes sample product, determine sample vector data
According to data label processing is grouped, acquisition group is to sample, according to group pair in sample vector data by corresponding data label
Sample, building obtain preset twin neural network model.
In one of the embodiments, group to sample include positive group to sample with negative group to sample；Model training module is also
Data label for being carried according to sample vector data classifies sample vector data, will be same according to permutation and combination
Two sample vector data of classification are grouped, and obtain positive group to sample, according to permutation and combination, by two different classes of samples
This vector data is grouped, and obtains negative group to sample.
Model construction module is also used to generate initial twin nerve net to sample according to group in one of the embodiments,
Network model obtains the model evaluation parameter of initial twin neural network model, when model evaluation parameter is not up to preset threshold model
When enclosing, it is adjusted by structural parameters of the back-propagation algorithm to initial twin neural network model, building obtains preset
Twin neural network model.
In one embodiment, early warning pushing module 500 is also used to determine risk according to product alert demand to be analyzed
Warning grade threshold value determines corresponding Risk-warning of product to be analyzed etc. according to risk probability and Risk-warning grade threshold
Grade, and push the Risk-warning information of product to be analyzed.
The said goods Risk-warning device is drawn a portrait using data to product to be analyzed by obtaining product information to be analyzed
Information is handled, and obtains the vector data for characterizing product to be analyzed, and vector data and sample vector data are carried out
Combination, inputs preset twin neural network model for combined result, obtains the risk probability of product to be analyzed, so that it is determined that
The Risk-warning grade of product is analyzed, the application is handled by data portrait, can deeply excavate the dependency number of product to be analyzed
According to being analysed to the vector data of product and sample data be combined as input data, and utilize preset twin nerve net
Network model evaluates the similarities of two inputs, makes the risk analysis result obtained and corresponding warning information is more accurate has
Effect.
Specific about product risks prior-warning device limits the limit that may refer to above for product risks method for early warning
Fixed, details are not described herein.Modules in the said goods Risk-warning device can fully or partially through software, hardware and its
Combination is to realize.Above-mentioned each module can be embedded in the form of hardware or independently of in the processor in computer equipment, can also be with
It is stored in the memory in computer equipment in a software form, in order to which processor calls the above modules of execution corresponding
Operation.
In one embodiment, a kind of computer equipment is provided, which can be terminal, internal structure
Figure can be as shown in Figure 5.The computer equipment includes processor, the memory, network interface, display connected by system bus
Screen and input unit.Wherein, the processor of the computer equipment is for providing calculating and control ability.The computer equipment is deposited
Reservoir includes non-volatile memory medium, built-in storage.The non-volatile memory medium is stored with operating system and computer journey
Sequence.The built-in storage provides environment for the operation of operating system and computer program in non-volatile memory medium.The calculating
The network interface of machine equipment is used to communicate with external terminal by network connection.When the computer program is executed by processor with
Realize a kind of product risks method for early warning.The display screen of the computer equipment can be liquid crystal display or electric ink is shown
Screen, the input unit of the computer equipment can be the touch layer covered on display screen, be also possible on computer equipment shell
Key, trace ball or the Trackpad of setting can also be external keyboard, Trackpad or mouse etc..
It will be understood by those skilled in the art that structure shown in Fig. 5, 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, the processor perform the steps of when executing computer program
Product information to be analyzed is obtained, the processing of data portrait is carried out to product information to be analyzed, obtains product letter to be analyzed
Cease corresponding vector data；
The sample vector data for obtaining preset twin neural network model, by vector data respectively with multiple sample vectors
Data are combined, and obtain multipair data splitting；
Multipair data splitting is inputted into preset twin neural network model, obtains the risk probability of product to be analyzed；
According to risk probability, the Risk-warning information of product to be analyzed is pushed.
In one embodiment, it is also performed the steps of when processor executes computer program
Each dimensional information of product to be analyzed is obtained, and according to default quantitative criteria, obtains the quantization number of each dimensional information
According to；
The processing of data portrait is carried out to the quantized data of each dimensional information, obtains the derivative data of product to be analyzed；
According to quantized data and derivative data, the vector data of product to be analyzed is obtained.
In one embodiment, it is also performed the steps of when processor executes computer program
Multipair data splitting is input to preset twin neural network model, obtains the highest data splitting of similarity；
According to the highest data splitting of similarity, sample vector data corresponding with product to be analyzed are obtained；
According to the risk situation of sample vector data corresponding product, the potential risk of product to be analyzed is obtained.
In one embodiment, it is also performed the steps of when processor executes computer program
Sample product is obtained, sample product is subjected to the processing of data portrait, obtains sample vector data；
Sample product is analyzed, determines the corresponding data label of sample vector data；
According to data label, processing is grouped in sample vector data, acquisition group is to sample；
According to group to sample, building obtains preset twin neural network model.
In one embodiment, group to sample include positive group to sample with negative group to sample；Processor executes computer journey
It is also performed the steps of when sequence
According to the data label that sample vector data carry, sample vector data are classified；
According to permutation and combination, same category of two sample vector data are grouped, obtain positive group to sample；
According to permutation and combination, two different classes of sample vector data are grouped, obtain negative group to sample.
In one embodiment, it is also performed the steps of when processor executes computer program
According to group to sample, initial twin neural network model is generated；
Obtain the model evaluation parameter of initial twin neural network model；
When model evaluation parameter is not up to preset threshold range, by back-propagation algorithm to initial twin neural network
The structural parameters of model are adjusted, and building obtains preset twin neural network model.
In one embodiment, it is also performed the steps of when processor executes computer program
According to product alert demand to be analyzed, Risk-warning grade threshold is determined；
According to risk probability and Risk-warning grade threshold, the corresponding Risk-warning grade of product to be analyzed is determined, and
Push the Risk-warning information of product to be analyzed.
The above-mentioned computer equipment for realizing product risks method for early warning is used by obtaining product information to be analyzed
Data portrait handles product information to be analyzed, obtains the vector data for characterizing product to be analyzed, and by vector number
It is combined according to sample vector data, combined result is inputted into preset twin neural network model, obtains product to be analyzed
Risk probability can deeply dig so that it is determined that the Risk-warning grade of product to be analyzed, the application are handled by data portrait
The related data for digging product to be analyzed, the vector data and sample data for being analysed to product are combined as input data, and
The similarity that two inputs are evaluated using preset twin neural network model makes the risk analysis result and correspondence obtained
Warning information it is more accurate effectively.
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
Product information to be analyzed is obtained, the processing of data portrait is carried out to product information to be analyzed, obtains product letter to be analyzed
Cease corresponding vector data；
The sample vector data for obtaining preset twin neural network model, by vector data respectively with multiple sample vectors
Data are combined, and obtain multipair data splitting；
Multipair data splitting is inputted into preset twin neural network model, obtains the risk probability of product to be analyzed；
According to risk probability, the Risk-warning information of product to be analyzed is pushed.
In one embodiment, it is also performed the steps of when computer program is executed by processor
Each dimensional information of product to be analyzed is obtained, and according to default quantitative criteria, obtains the quantization number of each dimensional information
According to；
The processing of data portrait is carried out to the quantized data of each dimensional information, obtains the derivative data of product to be analyzed；
According to quantized data and derivative data, the vector data of product to be analyzed is obtained.
In one embodiment, it is also performed the steps of when computer program is executed by processor
Multipair data splitting is input to preset twin neural network model, obtains the highest data splitting of similarity；
According to the highest data splitting of similarity, sample vector data corresponding with product to be analyzed are obtained；
According to the risk situation of sample vector data corresponding product, the potential risk of product to be analyzed is obtained.
In one embodiment, it is also performed the steps of when computer program is executed by processor
Sample product is obtained, sample product is subjected to the processing of data portrait, obtains sample vector data；
Sample product is analyzed, determines the corresponding data label of sample vector data；
According to data label, processing is grouped in sample vector data, acquisition group is to sample；
According to group to sample, building obtains preset twin neural network model.
In one embodiment, group to sample include positive group to sample with negative group to sample；Computer program is by processor
It is also performed the steps of when execution
According to the data label that sample vector data carry, sample vector data are classified；
According to permutation and combination, same category of two sample vector data are grouped, obtain positive group to sample；
According to permutation and combination, two different classes of sample vector data are grouped, obtain negative group to sample.
In one embodiment, it is also performed the steps of when computer program is executed by processor
According to group to sample, initial twin neural network model is generated；
Obtain the model evaluation parameter of initial twin neural network model；
When model evaluation parameter is not up to preset threshold range, by back-propagation algorithm to initial twin neural network
The structural parameters of model are adjusted, and building obtains preset twin neural network model.
In one embodiment, it is also performed the steps of when computer program is executed by processor
According to product alert demand to be analyzed, Risk-warning grade threshold is determined；
According to risk probability and Risk-warning grade threshold, the corresponding Risk-warning grade of product to be analyzed is determined, and
Push the Risk-warning information of product to be analyzed.
The above-mentioned computer readable storage medium for realizing product risks method for early warning is believed by obtaining product to be analyzed
Breath is handled product information to be analyzed using data portrait, obtains the vector data for characterizing product to be analyzed, and will
Vector data is combined with sample vector data, and combined result is inputted preset twin neural network model, is obtained wait divide
The risk probability of division product can be with so that it is determined that the Risk-warning grade of product to be analyzed, the application are handled by data portrait
The related data for deeply excavating product to be analyzed, is analysed to the vector data of product and sample data is combined as input number
According to, and evaluate using preset twin neural network model the similarities of two inputs, make the risk analysis result obtained with
And corresponding warning information is more accurate effectively.
Those of ordinary skill in the art will appreciate that realizing the whole in above-described embodiment product risks method for early warning or portion
Split flow is relevant hardware can be instructed to complete by computer program, computer program can be stored in one it is non-easily
In the property lost computer-readable storage medium, the computer program is when being executed, it may include such as the embodiment of above-mentioned each method
Process.Wherein, memory, storage, database or other media are appointed used in each embodiment provided herein
What is quoted, and may each comprise 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 storage
Device may include random access memory (RAM) or external cache.By way of illustration and not limitation, RAM is with a variety of
Form can obtain, such as static state RAM (SRAM), dynamic ram (DRAM), synchronous dram (SDRAM), double data rate sdram
(DDRSDRAM), enhanced 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.
Above embodiments only express the several embodiments of the application, and the description thereof is more specific and detailed, but can not
Therefore it is construed as limiting the scope of the patent.It should be pointed out that for those of ordinary skill in the art,
Under the premise of not departing from the application design, various modifications and improvements can be made, these belong to the protection scope of the application.
Therefore, the scope of protection shall be subject to the appended claims for the application patent.
Claims (10)
1. a kind of product risks method for early warning, which comprises
Product information to be analyzed is obtained, the processing of data portrait is carried out to the product information to be analyzed, obtains the production to be analyzed
The corresponding vector data of product information；
The sample vector data for obtaining preset twin neural network model, by the vector data respectively with multiple samples
Vector data is combined, and obtains multipair data splitting；
The multipair data splitting is inputted into the preset twin neural network model, the risk for obtaining product to be analyzed is general
Rate；
According to the risk probability, the Risk-warning information of the product to be analyzed is pushed.
2. product risks method for early warning according to claim 1, which is characterized in that it is described to obtain product information to be analyzed,
The processing of data portrait is carried out to the product information to be analyzed, obtains the corresponding vector data of the product information to be analyzed, packet
It includes:
Each dimensional information of product to be analyzed is obtained, and according to default quantitative criteria, obtains the quantization number of each dimensional information
According to；
The processing of data portrait is carried out to the quantized data of each dimensional information, obtains the derivative data of the product to be analyzed；
According to the quantized data and the derivative data, the vector data of the product to be analyzed is obtained.
3. product risks method for early warning according to claim 1, which is characterized in that further include:
The multipair data splitting is inputted into the preset twin neural network model, obtains the highest number of combinations of similarity
According to；
According to the highest data splitting of the similarity, sample vector data corresponding with the product to be analyzed are obtained；
According to the risk situation of the sample vector data corresponding product, the potential risk situation of the product to be analyzed is obtained.
4. product risks method for early warning according to claim 1, which is characterized in that described that the multipair data splitting is defeated
Enter the preset twin neural network model, before the risk probability for obtaining product to be analyzed, further includes:
Sample product is obtained, the sample product is subjected to the processing of data portrait, obtains sample vector data；
The sample product is analyzed, determines the corresponding data label of the sample vector data；
According to the data label, processing is grouped in the sample vector data, acquisition group is to sample；
According to described group to sample, building obtains preset twin neural network model.
5. product risks method for early warning according to claim 4, which is characterized in that described group includes positive group to sample to sample
This is with negative group to sample；It is described according to the data label, processing is grouped in the sample vector data, acquisition group is to sample
This, comprising:
According to the data label that sample vector data carry, sample vector data are classified；
According to permutation and combination, same category of two sample vector data are grouped, obtain positive group to sample；
According to permutation and combination, two different classes of sample vector data are grouped, obtain negative group to sample.
6. product risks method for early warning according to claim 4, which is characterized in that it is described according to described group to sample, structure
Build to obtain preset twin neural network model include:
According to described group to sample, initial twin neural network model is generated；
Obtain the model evaluation parameter of the initial twin neural network model；
When the model evaluation parameter is not up to preset threshold range, by back-propagation algorithm to the initial twin nerve
The structural parameters of network model are adjusted, and building obtains the preset twin neural network model.
7. product risks method for early warning according to claim 1, which is characterized in that it is described according to the risk probability, it pushes away
Send the Risk-warning information of the product to be analyzed, comprising:
According to product alert demand to be analyzed, Risk-warning grade threshold is determined；
According to the risk probability and the Risk-warning grade threshold, the corresponding Risk-warning of the product to be analyzed is determined
Grade, and push the Risk-warning information of the product to be analyzed.
8. a kind of product risks prior-warning device, which is characterized in that described device includes:
Data processing module carries out the processing of data portrait to the product information to be analyzed for obtaining product information to be analyzed,
Obtain the corresponding vector data of the product information to be analyzed；
Data combination module obtains multipair group for the vector data to be combined with multiple sample vector data respectively
Close data；
Data analysis module obtains to be analyzed for the multipair data splitting to be inputted preset twin neural network model
The risk probability of product；
Risk-warning module, for pushing the Risk-warning information of the product to be analyzed according to the risk probability.
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 7 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 7 is realized when being executed by processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811089567.5A CN109360105A (en) | 2018-09-18 | 2018-09-18 | Product risks method for early warning, device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811089567.5A CN109360105A (en) | 2018-09-18 | 2018-09-18 | Product risks method for early warning, device, computer equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109360105A true CN109360105A (en) | 2019-02-19 |
Family
ID=65351252
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811089567.5A Pending CN109360105A (en) | 2018-09-18 | 2018-09-18 | Product risks method for early warning, device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109360105A (en) |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110097250A (en) * | 2019-03-20 | 2019-08-06 | 平安直通咨询有限公司上海分公司 | Product risks prediction technique, device, computer equipment and storage medium |
| CN110135694A (en) * | 2019-04-12 | 2019-08-16 | 深圳壹账通智能科技有限公司 | Product risks appraisal procedure, device, computer equipment and storage medium |
| CN110288468A (en) * | 2019-04-19 | 2019-09-27 | 平安科技（深圳）有限公司 | Data characteristics method for digging, device, electronic equipment and storage medium |
| CN110378780A (en) * | 2019-06-17 | 2019-10-25 | 深圳壹账通智能科技有限公司 | Air control strategy adjusting method, device, computer equipment and storage medium |
| CN110490720A (en) * | 2019-07-08 | 2019-11-22 | 深圳壹账通智能科技有限公司 | Financial data analysis and early warning method, apparatus, computer equipment and storage medium |
| CN111779541A (en) * | 2020-06-08 | 2020-10-16 | 云南昆钢电子信息科技有限公司 | Method and device for monitoring and early warning of underground environment |
| CN111798322A (en) * | 2020-06-30 | 2020-10-20 | 中国工商银行股份有限公司 | Data processing method and device for bank financing product |
| CN112115374A (en) * | 2019-06-20 | 2020-12-22 | 腾讯科技（深圳）有限公司 | Relationship chain portrait method, device, storage medium and computer equipment |
| CN112580916A (en) * | 2019-09-30 | 2021-03-30 | 深圳无域科技技术有限公司 | Data evaluation method and device, computer equipment and storage medium |
| CN112749326A (en) * | 2019-11-15 | 2021-05-04 | 腾讯科技（深圳）有限公司 | Information processing method, information processing device, computer equipment and storage medium |
| CN112884204A (en) * | 2021-01-22 | 2021-06-01 | 中国科学院信息工程研究所 | Network security risk event prediction method and device |
| CN112989784A (en) * | 2021-03-04 | 2021-06-18 | 广州汇才创智科技有限公司 | Text automatic scoring method and device based on twin neural network and electronic equipment |
| CN113297338A (en) * | 2021-07-27 | 2021-08-24 | 平安科技（深圳）有限公司 | Method, device and equipment for generating product recommendation path and storage medium |
| CN113342846A (en) * | 2021-06-29 | 2021-09-03 | 福州外语外贸学院 | Big data-based early warning and control reminding method and device and computer equipment |
| CN113450116A (en) * | 2020-03-24 | 2021-09-28 | 国家计算机网络与信息安全管理中心 | Transaction risk analysis method and device, electronic equipment and storage medium |
| CN113723811A (en) * | 2021-08-30 | 2021-11-30 | 平安国际智慧城市科技股份有限公司 | Equipment maintenance unit evaluation method and device based on machine learning and electronic equipment |
| CN113780827A (en) * | 2021-09-14 | 2021-12-10 | 北京沃东天骏信息技术有限公司 | Article screening method and device, electronic equipment and computer readable medium |
| CN114785616A (en) * | 2022-06-13 | 2022-07-22 | 深圳红途科技有限公司 | Data risk detection method and device, computer equipment and storage medium |
| CN116977122A (en) * | 2023-07-06 | 2023-10-31 | 双龙软创(深圳)科技有限公司 | Remote automatic monitoring method for dangerous rooms based on digital twin technology |
| CN117237114A (en) * | 2023-11-10 | 2023-12-15 | 深圳市迪博企业风险管理技术有限公司 | Financing trade compliance detection method based on twin evolution |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106779755A (en) * | 2016-12-31 | 2017-05-31 | 湖南文沥征信数据服务有限公司 | A kind of network electric business borrows or lends money methods of risk assessment and model |
| CN107248030A (en) * | 2017-05-26 | 2017-10-13 | 谢首鹏 | A kind of bond Risk Forecast Method and system based on machine learning algorithm |
| CN107729143A (en) * | 2017-09-30 | 2018-02-23 | 广东欧珀移动通信有限公司 | Application control method, apparatus, storage medium and electronic equipment |
-
2018
- 2018-09-18 CN CN201811089567.5A patent/CN109360105A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106779755A (en) * | 2016-12-31 | 2017-05-31 | 湖南文沥征信数据服务有限公司 | A kind of network electric business borrows or lends money methods of risk assessment and model |
| CN107248030A (en) * | 2017-05-26 | 2017-10-13 | 谢首鹏 | A kind of bond Risk Forecast Method and system based on machine learning algorithm |
| CN107729143A (en) * | 2017-09-30 | 2018-02-23 | 广东欧珀移动通信有限公司 | Application control method, apparatus, storage medium and electronic equipment |
Non-Patent Citations (2)
| Title |
|---|
| FIRDAOUSS DOUKKALI: ""单样本学习：使用孪生神经网络进行人脸识别"", pages 1 - 2, Retrieved from the Internet <URL:https://cloud.tencent.com/developer/article/1016699> * |
| 雷晓静: ""基于深度学习的图像分类研究"", 《中国优秀硕士论文全文数据库信息科技辑（月刊）》, no. 04, 15 April 2018 (2018-04-15), pages 37 - 44 * |
Cited By (28)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110097250A (en) * | 2019-03-20 | 2019-08-06 | 平安直通咨询有限公司上海分公司 | Product risks prediction technique, device, computer equipment and storage medium |
| CN110135694A (en) * | 2019-04-12 | 2019-08-16 | 深圳壹账通智能科技有限公司 | Product risks appraisal procedure, device, computer equipment and storage medium |
| CN110288468A (en) * | 2019-04-19 | 2019-09-27 | 平安科技（深圳）有限公司 | Data characteristics method for digging, device, electronic equipment and storage medium |
| CN110288468B (en) * | 2019-04-19 | 2023-06-06 | 平安科技（深圳）有限公司 | Data feature mining method and device, electronic equipment and storage medium |
| CN110378780A (en) * | 2019-06-17 | 2019-10-25 | 深圳壹账通智能科技有限公司 | Air control strategy adjusting method, device, computer equipment and storage medium |
| CN112115374B (en) * | 2019-06-20 | 2023-10-27 | 腾讯科技（深圳）有限公司 | Relationship chain representation method, device, storage medium and computer equipment |
| CN112115374A (en) * | 2019-06-20 | 2020-12-22 | 腾讯科技（深圳）有限公司 | Relationship chain portrait method, device, storage medium and computer equipment |
| CN110490720A (en) * | 2019-07-08 | 2019-11-22 | 深圳壹账通智能科技有限公司 | Financial data analysis and early warning method, apparatus, computer equipment and storage medium |
| CN112580916B (en) * | 2019-09-30 | 2024-05-28 | 深圳无域科技技术有限公司 | Data evaluation method, device, computer equipment and storage medium |
| CN112580916A (en) * | 2019-09-30 | 2021-03-30 | 深圳无域科技技术有限公司 | Data evaluation method and device, computer equipment and storage medium |
| CN112749326A (en) * | 2019-11-15 | 2021-05-04 | 腾讯科技（深圳）有限公司 | Information processing method, information processing device, computer equipment and storage medium |
| CN112749326B (en) * | 2019-11-15 | 2023-10-03 | 腾讯科技（深圳）有限公司 | Information processing method, information processing device, computer equipment and storage medium |
| CN113450116A (en) * | 2020-03-24 | 2021-09-28 | 国家计算机网络与信息安全管理中心 | Transaction risk analysis method and device, electronic equipment and storage medium |
| CN111779541A (en) * | 2020-06-08 | 2020-10-16 | 云南昆钢电子信息科技有限公司 | Method and device for monitoring and early warning of underground environment |
| CN111798322A (en) * | 2020-06-30 | 2020-10-20 | 中国工商银行股份有限公司 | Data processing method and device for bank financing product |
| CN112884204A (en) * | 2021-01-22 | 2021-06-01 | 中国科学院信息工程研究所 | Network security risk event prediction method and device |
| CN112884204B (en) * | 2021-01-22 | 2024-04-12 | 中国科学院信息工程研究所 | Network security risk event prediction method and device |
| CN112989784A (en) * | 2021-03-04 | 2021-06-18 | 广州汇才创智科技有限公司 | Text automatic scoring method and device based on twin neural network and electronic equipment |
| CN113342846B (en) * | 2021-06-29 | 2024-03-19 | 福州外语外贸学院 | Early warning handle control reminding method and device based on big data and computer equipment |
| CN113342846A (en) * | 2021-06-29 | 2021-09-03 | 福州外语外贸学院 | Big data-based early warning and control reminding method and device and computer equipment |
| CN113297338A (en) * | 2021-07-27 | 2021-08-24 | 平安科技（深圳）有限公司 | Method, device and equipment for generating product recommendation path and storage medium |
| CN113723811A (en) * | 2021-08-30 | 2021-11-30 | 平安国际智慧城市科技股份有限公司 | Equipment maintenance unit evaluation method and device based on machine learning and electronic equipment |
| CN113780827A (en) * | 2021-09-14 | 2021-12-10 | 北京沃东天骏信息技术有限公司 | Article screening method and device, electronic equipment and computer readable medium |
| CN114785616A (en) * | 2022-06-13 | 2022-07-22 | 深圳红途科技有限公司 | Data risk detection method and device, computer equipment and storage medium |
| CN116977122A (en) * | 2023-07-06 | 2023-10-31 | 双龙软创(深圳)科技有限公司 | Remote automatic monitoring method for dangerous rooms based on digital twin technology |
| CN116977122B (en) * | 2023-07-06 | 2024-04-19 | 双龙软创(深圳)科技有限公司 | Remote automatic monitoring method for dangerous rooms based on digital twin technology |
| CN117237114B (en) * | 2023-11-10 | 2024-03-08 | 深圳市迪博企业风险管理技术有限公司 | Financing trade compliance detection method based on twin evolution |
| CN117237114A (en) * | 2023-11-10 | 2023-12-15 | 深圳市迪博企业风险管理技术有限公司 | Financing trade compliance detection method based on twin evolution |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109360105A (en) | Product risks method for early warning, device, computer equipment and storage medium | |
| Zhou et al. | Stock market prediction on high-frequency data using generative adversarial nets. | |
| Souma et al. | Enhanced news sentiment analysis using deep learning methods | |
| Ma et al. | Financial credit risk prediction in internet finance driven by machine learning | |
| Park et al. | Stock market forecasting using a multi-task approach integrating long short-term memory and the random forest framework | |
| Wu et al. | The analysis of credit risks in agricultural supply chain finance assessment model based on genetic algorithm and backpropagation neural network | |
| Wang et al. | Software defect prediction model based on LASSO–SVM | |
| CN108491406B (en) | Information classification method and device, computer equipment and storage medium | |
| Trigueiros et al. | Neural networks and empirical research in accounting | |
| CN108898476A (en) | A kind of loan customer credit-graded approach and device | |
| CN111126668A (en) | Spark operation time prediction method and device based on graph convolution network | |
| CN110852881B (en) | Risk account identification method and device, electronic equipment and medium | |
| Kang | Financial risk assessment model based on big data | |
| Zhou et al. | Numerical solution for ruin probability of continuous time model based on neural network algorithm | |
| CN113674087A (en) | Enterprise credit rating method, apparatus, electronic device and medium | |
| CN110442713A (en) | Abstract generation method, apparatus, computer equipment and storage medium | |
| Wimmer et al. | Leveraging vision-language models for granular market change prediction | |
| CN110310012A (en) | Data analysing method, device, equipment and computer readable storage medium | |
| Peng et al. | Attention-based CNN–LSTM for high-frequency multiple cryptocurrency trend prediction | |
| Zhong et al. | Construction project risk prediction model based on EW-FAHP and one dimensional convolution neural network | |
| Unadkat et al. | Deep learning for financial prediction | |
| CN110766465A (en) | Financial product evaluation method and verification method and device thereof | |
| CN110909975A (en) | Scientific research platform benefit evaluation method and device | |
| Yao et al. | Six-factor asset pricing and portfolio investment via deep learning: Evidence from Chinese stock market | |
| CN109829745A (en) | Business revenue data predication method, device, computer equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |