Risk appraisal procedure and device
Technical field
The invention belongs to risk assessment technology field, more particularly to a kind of risk appraisal procedure and device.
Background technology
Risk assessment is by being analyzed to the ill risk indicator of user, determining the mistake of the ill risk of user
Journey.Because the risk indicator for affecting risk assessment has various, and each risk indicator to the influence degree of risk not
It is identical so that risk assessment has extremely strong uncertainty.Bayesian network has extremely strong reasoning and cause diagnosis work(
Can, it is one of common method of risk assessment.Bayesian network is applied when risk is assessed, it is impossible to directly judge illness
Risk assessment value, but the mutual relation of each risk indicator can be identified in complication system, it is a kind of all wind of overall consideration
Nearly refer to calibration method, can may find association risk indicator the strongest in risk indicator.Existing utilization Bayesian network
When carrying out risk assessment, generally risk indicator wind is not accounted for into directly as the factor for evaluating final risk value
Being overlapped mutually between dangerous index and offset, the problem for causing the accuracy of risk value for obtaining low.
Therefore, it is necessary to a kind of new technical scheme is proposed, to solve above-mentioned technical problem.
The content of the invention
In consideration of it, the embodiment of the present invention provides a kind of risk appraisal procedure and device, to solve existing utilization pattra leaves
When this network carries out risk assessment, generally risk indicator is not had directly as the factor for evaluating final risk value
Consider being overlapped mutually between risk indicator and offset, the problem for causing the accuracy of the risk value of acquisition low.
A kind of first aspect of the embodiment of the present invention, there is provided risk appraisal procedure, including:
Obtain the ill target risk index score value of targeted customer;
Based on the target risk index score value, corresponding target risk classification score value is determined;
Based on the target risk classification score value, the target risk assessment models for training are called to enter the targeted customer
Row risk is assessed;The target risk assessment models include:Target risk value=∑ target risk classification score value * wind
Dangerous class weight, the target risk classification score value=∑ target risk index score value * risk indicator prevalences;
Obtain the ill target risk value of the targeted customer.
A kind of second aspect of the embodiment of the present invention, there is provided risk apparatus for evaluating, including:
First acquisition module, the target risk index score value ill for obtaining targeted customer;
Determining module, based on the target risk index score value, determines corresponding target risk classification score value;
Calling module, based on the target risk classification score value, calls the target risk assessment models for training to described
Targeted customer carries out risk assessment;The target risk assessment models include:Target risk value=∑ target risk
Classification score value * kind of risk weights, the target risk classification score value=∑ target risk index score value * risk indicators are ill
Rate;
Second acquisition module, the target risk value ill for obtaining the targeted customer.
The beneficial effect that the embodiment of the present invention is present compared with prior art is:The method or apparatus of the embodiment of the present invention
In, obtain the ill target risk index score value of targeted customer and determine corresponding target risk classification score value, recall training
Good target risk assessment models carry out risk assessment to targeted customer, to obtain target risk value.Suffer from target
During the determination of sick value-at-risk, it is considered to different target kind of risk score value and corresponding kind of risk weight, also, each
During the determination of target risk classification score value, it is considered to different target risk indicator score value and corresponding risk indicator prevalence,
So as to realize being overlapped mutually and offset between different target risk indicator, to improve the accuracy of target risk value.And,
Risk evaluation process is carried out to targeted customer using the target risk assessment models for training, calculates simple and convenient, processed
Efficiency is very fast.
Description of the drawings
Technical scheme in order to be illustrated more clearly that the embodiment of the present invention, below will be to embodiment or description of the prior art
Needed for the accompanying drawing to be used be briefly described, it should be apparent that, drawings in the following description be only the present invention some
Embodiment, for those of ordinary skill in the art, without having to pay creative labor, can be with according to these
Accompanying drawing obtains other accompanying drawings.
Fig. 1 is the flowchart one of the risk appraisal procedure that the embodiment of the present invention one is provided;
Fig. 2 is the flowchart two of the risk appraisal procedure that the embodiment of the present invention one is provided;
Fig. 3 is the flowchart three of the risk appraisal procedure that the embodiment of the present invention one is provided;
Fig. 4 is the structured flowchart of the risk apparatus for evaluating that the embodiment of the present invention two is provided.
Specific embodiment
In order that the objects, technical solutions and advantages of the present invention become more apparent, it is right below in conjunction with drawings and Examples
The present invention is described in detail.It should be appreciated that specific embodiment described herein is not used to only to explain the present invention
Limit the present invention.
The embodiment of the present invention provides a kind of risk appraisal procedure.In order to illustrate that risk provided by the present invention is commented
Estimate method, illustrate below by specific embodiment.
Embodiment one
Fig. 1 shows the flowchart one of the risk appraisal procedure that the embodiment of the present invention one is provided.The ill wind
The step of dangerous appraisal procedure, details are as follows:
S101:Obtain the ill target risk index score value of targeted customer.
Wherein, targeted customer refers to be currently needed for assessing and suffers from certain sick user, and target risk index is included but is not limited to
Age risk indicator, sex risk indicator, Ethanol intake amount risk indicator, smoking risk indicator, triglyceride risk indicator, gallbladder
Sterin risk indicator, hyperglycemia risk indicator, high density lipoprotein risk indicator and family illness history risk indicator.Target risk
Index score value refers to the ratio for accounting for all target risk indexs of a certain item target risk indicator in the target risk index.Obtain
The ill target risk index score value of targeted customer is taken equivalent to the weight for getting each target risk index, i.e., each target risk
Significance level of the index when affecting targeted customer ill, it is to avoid after the objectives risk indicator score value is defaulted as identical causing
Result inaccurate situation during continuous assessment risk.
S102:Based on the target risk index score value, corresponding target risk classification score value, the target risk are determined
Classification score value=∑ target risk index score value * risk indicator prevalences.
According to default classification criteria, target risk index is summarized as into several target risk classifications.Obtained according to data base
The corresponding risk indicator prevalence of risk indicator, then the target risk index score value obtained in S101, by formula:Target
Kind of risk score value=∑ target risk index score value * risk indicators prevalence is calculated and obtains each target risk classification score value.
The step is sorted out target risk index, it is contemplated that being overlapped mutually between target risk index and supported
Disappear, the accuracy of result when increased further evaluation risk.
S103:Based on the target risk classification score value, the target risk assessment models for training are called to the target
User carries out risk assessment;The target risk assessment models include:Target risk value=∑ target risk classification
Score value * kind of risk weights, the target risk classification score value=∑ target risk index score value * risk indicator prevalences.
Kind of risk weight according to target, it is known that according to the target risk classification score value calculated in S102, suffer from
Sick value-at-risk=∑ target risk classification score value * kind of risk weight carries out risk assessment to the targeted customer.
S104:Obtain the ill target risk value of the targeted customer.
Target risk value is a probability, obtains the ill probability of targeted customer.The probability is only as a reference
Assessed value.
The ill target risk index score value of targeted customer is obtained in the embodiment of the present invention and corresponding target risk is determined
Classification score value, recall the target risk assessment models for training carries out risk assessment to targeted customer, to obtain target
Risk value.During the determination of target risk value, it is considered to different target kind of risk score value and corresponding risk
Class weight, also, during the determination of each target risk classification score value, it is considered to different target risk indicator score value and phase
The risk indicator prevalence answered, it is ill to improve target so as to realize being overlapped mutually and offset between different target risk indicator
The accuracy of value-at-risk.And, risk is carried out to targeted customer using the target risk assessment models for training and was assessed
Journey, calculates simple and convenient, and treatment effeciency is very fast.
Fig. 2 shows the flowchart two of the risk appraisal procedure that the embodiment of the present invention one is provided.
Further, risk appraisal procedure also includes:The target risk assessment models that acquisition is trained, the acquisition
The target risk assessment models for training include:
S201:Obtain training user and concentrate the training risk indicator score value of each training user, and determine each training wind
The corresponding risk indicator prevalence of dangerous index score value.
Wherein, each training user to training user to concentrate is trained, and obtains training user and concentrates each training to use
The training risk indicator score value at family, according to data base the corresponding risk indicator prevalence of each training risk indicator score value is determined.
S202:Based on the training risk indicator score value and the risk indicator prevalence, training kind of risk point is obtained
Value;The training kind of risk score value=∑ training risk indicator score value * risk indicator prevalences.
According to default classification criteria, target risk index is summarized as into several target risk classifications.Train according in S201
Risk indicator score value and risk indicator prevalence, training risk indicator score value is multiplied and is added up with risk indicator prevalence to be obtained
Training kind of risk score value.
S203:Obtain the original risk value that the training user concentrates all training users.
The original risk value that the training user concentrates all training users, original ill wind are obtained according to data base
Danger value is equivalent to prior probability.
S204:Initial risks assessment models are built, the initial risks assessment models include:Training risk assessment value=∑
Training kind of risk score value * kind of risk weights.
Wherein, kind of risk score value is trained, it is known that training risk assessment value to obtain, kind of risk Decision Makings with Weights Unknown builds
Kind of risk weight is obtained by only need to being trained to the training user that training user concentrates after the model.
S205:Using the training kind of risk score value and original risk value, to the initial risks assessment models
It is trained, determines the kind of risk weight, obtains the target risk assessment models for training.
According to the training kind of risk score value for having obtained and original risk value, adopting includes EM algorithms to initially commenting
Estimate model (training risk assessment value=∑ training kind of risk score value * kind of risk weights) to be trained, determine kind of risk
Weight.The training process is carried out a progressive alternate and is estimated based on the sample (only partial risks index) of shortage of data
Process, until obtaining the kind of risk weight of optimum or local optimum.The step is conducive to carrying out ill wind to targeted customer
During the assessment of danger, directly its weight is obtained according to the title of kind of risk correspondence.
In the present embodiment, it is trained by each training user to training user to concentrate, obtains training kind of risk
Score value, builds initial risks assessment models and determines kind of risk weight, training user's collection substantial amounts, and what is obtained after training comments
There is versatility when estimating model for the risk for assessing targeted customer and accuracy is high.
Fig. 3 shows the flowchart three of the risk appraisal procedure that the embodiment of the present invention one is provided.
Preferably, after the ill target risk value of the targeted customer is obtained, also include:
S301:Using risk Logarithm conversion formula, the target risk value is converted into into risk log-rank.
Conversion formula is as follows:Pl=lgP+5.Wherein, Pl represents logarithm risk class, and P is represented according to the risk
The risk that assessment models are evaluated.
Risk log-rank after conversion can intuitively reflect the acceptable degree of specific risk, for doctor enters
Row diagnosis decision making process more has directive significance.
S302:Statistical analysiss are carried out based on the risk log-rank.
Substantial amounts of targeted customer is carried out after risk assessment, then changed according to the target user data for updating
In generation, storehouse is updated the data according to statistical analysiss, at the same with new risk class weight, so that cycle analyses are used, so can be gradually
The value of kind of risk weight is improved, increases the accuracy of risk assessment.
Preferably, the structure initial risks assessment models, including:Initial risks are built based on bayesian network structure to comment
Estimate model, the bayesian network structure includes risk, three layers of oriented structure of kind of risk and risk indicator.
Preferably, the kind of risk refers to including Constitution Indexes classification, behavioral indicator classification, health check-up index classification and heredity
Mark classification;The Constitution Indexes classification includes age risk indicator and sex risk indicator, and the behavioral indicator classification includes wine
Smart intake risk indicator and smoking risk indicator, the health check-up index classification includes triglyceride risk indicator, cholesterol wind
Dangerous index, hyperglycemia risk indicator and high density lipoprotein risk indicator, the Heredity index classification includes family's illness history wind
Dangerous index.
Embodiment two
Corresponding to the risk appraisal procedure described in foregoing embodiments, Fig. 4 shows trouble provided in an embodiment of the present invention
The structured flowchart of sick risk assessment device.
With reference to Fig. 4, the risk apparatus for evaluating includes:
First acquisition module 41, the target risk index score value ill for obtaining targeted customer.
Determining module 42, based on the target risk index score value, determines corresponding target risk classification score value.
Calling module 43, based on the target risk classification score value, calls the target risk assessment models for training to institute
Stating targeted customer carries out risk assessment;The target risk assessment models include:Target risk value=∑ target wind
Dangerous classification score value * kind of risk weights, the target risk classification score value=∑ target risk index score value * risk indicators are ill
Rate.
Second acquisition module 44, the target risk value ill for obtaining the targeted customer.
Alternatively, described device also includes:3rd acquisition module, for obtaining the target risk assessment models for training,
3rd acquisition module includes:
First acquisition unit, for the training risk indicator score value that acquisition training user concentrates each training user, and really
Determine the corresponding risk indicator prevalence of each training risk indicator score value.
Second acquisition unit, for based on the training risk indicator score value and the risk indicator prevalence, obtaining instruction
Practice kind of risk score value;The training kind of risk score value=∑ training risk indicator score value * risk indicator prevalences.
3rd acquiring unit, for obtaining the original risk value that the training user concentrates all training users.
Construction unit, for building initial risks assessment models, the initial risks assessment models include:Training risk is commented
Valuation=∑ training kind of risk score value * kind of risk weights.
The initial risks, using the training kind of risk score value and original risk value, are assessed by training unit
Model is trained, and determines the kind of risk weight, to obtain the target risk assessment models for training.
Alternatively, the structure initial risks assessment models, including:Initial risks are built based on bayesian network structure to comment
Estimate model, the bayesian network structure includes risk, three layers of oriented structure of kind of risk and risk indicator.
Alternatively, the kind of risk refers to including Constitution Indexes classification, behavioral indicator classification, health check-up index classification and heredity
Mark classification;The Constitution Indexes classification includes age risk indicator and sex risk indicator, and the behavioral indicator classification includes wine
Smart intake risk indicator and smoking risk indicator, the health check-up index classification includes triglyceride risk indicator, cholesterol wind
Dangerous index, hyperglycemia risk indicator and high density lipoprotein risk indicator, the Heredity index classification includes family's illness history wind
Dangerous index.
Alternatively, described device also includes:
Modular converter, using risk Logarithm conversion formula, by the target risk value risk log-rank is converted into.
Statistical module, based on the risk log-rank statistical analysiss are carried out.
In embodiment provided by the present invention, it should be understood that disclosed apparatus and method, can pass through other
Mode is realized.For example, system embodiment described above is only schematic, for example, the division of the unit or unit,
It is only a kind of division of logic function, there can be other dividing mode when actually realizing, such as multiple units or component can be with
With reference to or be desirably integrated into another system, or some features can be ignored, or not perform.It is another, it is shown or discussed
Coupling each other or direct-coupling or communication connection can be INDIRECT COUPLING by some interfaces, device or unit or
Communication connection, can be electrical, mechanical or other forms.
The unit as separating component explanation can be or may not be it is physically separate, it is aobvious as unit
The part for showing can be or may not be physical location, you can with positioned at a place, or can also be distributed to multiple
On NE.Some or all of unit therein can according to the actual needs be selected to realize the mesh of this embodiment scheme
's.
In addition, each functional unit in each embodiment of the invention can be integrated in a processing unit, it is also possible to
It is that unit is individually physically present, it is also possible to which two or more units are integrated in a unit.Above-mentioned integrated list
Unit both can be realized in the form of hardware, it would however also be possible to employ the form of SFU software functional unit is realized.
If the integrated unit is realized using in the form of SFU software functional unit and as independent production marketing or used
When, during a computer read/write memory medium can be stored in.Based on such understanding, the technical scheme of the embodiment of the present invention
The part for substantially contributing to prior art in other words or all or part of the technical scheme can be with software products
Form embody, the computer software product is stored in a storage medium, including some instructions use so that one
Computer equipment (can be personal computer, server, or network equipment etc.) or processor (processor) perform this
The all or part of step of bright embodiment each embodiment methods described.And aforesaid storage medium includes:USB flash disk, portable hard drive,
Read only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic
Dish or CD etc. are various can be with the medium of store program codes.
Embodiment described above only to illustrate technical scheme, rather than a limitation;Although with reference to aforementioned reality
Apply example to be described in detail the present invention, it will be understood by those within the art that:It still can be to aforementioned each
Technical scheme described in embodiment is modified, or carries out equivalent to which part technical characteristic;And these are changed
Or replace, the spirit and scope of the essence disengaging various embodiments of the present invention technical scheme of appropriate technical solution are not made, all should
It is included within protection scope of the present invention.