CN109978406A - A kind of method and system of security downside risks assessment diagnosis - Google Patents
A kind of method and system of security downside risks assessment diagnosis Download PDFInfo
- Publication number
- CN109978406A CN109978406A CN201910277134.0A CN201910277134A CN109978406A CN 109978406 A CN109978406 A CN 109978406A CN 201910277134 A CN201910277134 A CN 201910277134A CN 109978406 A CN109978406 A CN 109978406A
- Authority
- CN
- China
- Prior art keywords
- variable
- security
- model
- final mask
- dynamic
- 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 50
- 238000003745 diagnosis Methods 0.000 title claims abstract description 35
- 238000012549 training Methods 0.000 claims description 19
- 238000011156 evaluation Methods 0.000 claims description 15
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000012360 testing method Methods 0.000 claims description 5
- 238000005192 partition Methods 0.000 claims description 4
- 238000012216 screening Methods 0.000 claims description 4
- 238000012163 sequencing technique Methods 0.000 claims description 4
- 238000012502 risk assessment Methods 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 5
- 241000039077 Copula Species 0.000 description 4
- PEDCQBHIVMGVHV-UHFFFAOYSA-N Glycerine Chemical compound OCC(O)CO PEDCQBHIVMGVHV-UHFFFAOYSA-N 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 230000008859 change Effects 0.000 description 2
- 238000007477 logistic regression Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 238000013480 data collection Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000004744 fabric Substances 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005498 polishing Methods 0.000 description 1
- 238000013441 quality evaluation Methods 0.000 description 1
- 238000000714 time series forecasting 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
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- 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
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Game Theory and Decision Science (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Operations Research (AREA)
- General Physics & Mathematics (AREA)
- Technology Law (AREA)
- Educational Administration (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of method and system of security downside risks assessment diagnosis, the method for security downside risks assessment diagnosis provided by the invention, including data set is constructed according to target variable and quantitative characteristic variable;According to collection selection dynamic notable feature, model variable is obtained, establishes dynamic scoring model, code of points is generated according to dynamic scoring model, obtains final mask score value;Low-risk probability value and final probability distribution value are calculated according to final mask score value, obtains standard security point, according to standard security point, obtains diagnostic text.The present invention by providing a kind of method and system of security downside risks assessment diagnosis, improve it is following to security downside risks and portfolio can the assessment accuracy that carries out of the loss of energy.
Description
Technical field
The present invention relates to securities risk assessment technology field more particularly to a kind of methods of security downside risks assessment diagnosis
And system.
Background technique
The method for being mainly based upon time series forecasting for the method for assessing security downside risks at present, this method are claimed
Make the Time-varying Copula estimation technique.
The Time-varying Copula estimation technique is mainly to predict the historical price time series data of assessment security, obtains future
The price distribution of a period of time obtains Time-varying Copula estimation technique value to calculate.Since the Time-varying Copula estimation technique is based on pure
Security price time series predicting model, be unable to integrated multidimensional degree information and with reference to similar card Securities price data change make more
Accurately comprehensive judgement, especially to the risk assessment as unit of day or the moon, error can be become very large.
Summary of the invention
In order to solve problems in the prior art, the embodiment of the invention provides a kind of sides of security downside risks assessment diagnosis
Method and system.The technical solution is as follows:
In a first aspect, providing a kind of method of security downside risks assessment diagnosis, which comprises
Data set is constructed according to target variable and quantitative characteristic variable;
According to the collection selection dynamic notable feature, model variable is obtained, dynamic is established according to the model variable
Scoring model generates code of points according to the dynamic scoring model, obtains final mask score value according to the code of points;
Low-risk probability value and final probability distribution value are calculated according to the final mask score value, obtains standard security
Point, according to the standard security point, obtain diagnostic text.
Optionally, data set is constructed according to target variable and quantitative characteristic variable, comprising:
It is described according to the feature of acquisition, by quantitatively portraying, obtains the quantitative characteristic variable;
It is withdrawn according to maximum with descending ranking, obtains the target variable;
The data set is constructed according to the time sequencing of the quantitative characteristic variable and the target variable;
Training set and test set are divided into according to the data set constructed.
Optionally, according to the collection selection dynamic notable feature, model variable is obtained, is built according to the model variable
Vertical dynamic scoring model generates code of points according to the dynamic scoring model, obtains final mask according to the code of points
Score value, comprising:
The quantitative characteristic variable in the training set is chosen, evidence weight is calculated in input data training pattern
Value and information magnitude;
Branch mailbox and screening to the quantitative characteristic variable, output are carried out according to the value of the quantitative characteristic variable
Model variable;
According to the calculated related coefficient of the model variable, the model variable is further screened, exports final mould
Type variable obtains having the associated final mask variable of conspicuousness to the target variable;
According to the model variable in conjunction with the target variable, regression model is constructed, obtains fit equation;
The training set is fitted according to the fit equation, obtains branch mailbox property parameters and intercept, is weighed by adjusting
Formula updates variable's attribute coefficient, and the variable's attribute coefficient exports in the regression model according to the variation of the data set
The dynamic grading rule;
According to the branch mailbox property parameters and the intercept, is calculated by original point of formula of security, export the final mould
Type scoring.
Optionally, low-risk probability value and final probability distribution value are calculated according to the final mask score value, is marked
Quasi- safety point obtains diagnostic text according to the standard security point, comprising:
Low-risk probability value is obtained according to scorecard odds principle;
The low-risk probability value is distinguished into two kinds of differences of input linear interpolating estimation method and partition balancing evaluation method
Evaluation method, export the low-risk probability value of two kinds of evaluation methods respectively, take out average value, obtain risk ranking probability point
Cloth;
It obtains target security maximum in different time periods and withdraws rate, divide calculation formula according to standard security, obtain the mesh
The standard security of standard card certificate point;
According to the score value of the dynamic grading Rule security items distinguished variable, the score value is carried out pair
Than being equipped with evaluation text to every distinguished variable in advance, finding corresponding evaluation text according to the standard security point
Obtain assessment diagnostic text.
Optionally, described further to screen the model variable, final mask variable is exported, obtains becoming the target
Measurer has the associated final mask variable of conspicuousness, comprising:
The model variable input initialization model variable is concentrated, the final mask variables set of output;
Choose the candidate model variable;
Determine the final mask variable;
Determine that the model variable in the final mask variables set is the final mask variable.
Second aspect provides a kind of system of security downside risks assessment diagnosis, for executing above-mentioned security descending airflow
The method of danger assessment diagnosis.
The third aspect provides a kind of computer, and the computer includes processor and memory, deposits in the memory
Contain at least one instruction, at least a Duan Chengxu, code set or instruction set, at least one instruction, an at least Duan Cheng
The method that sequence, the code set or instruction set are loaded by the processor and executed to realize the assessment diagnosis of security downside risks.
Fourth aspect provides a kind of computer readable storage medium, at least one finger is stored in the storage medium
Enable, at least a Duan Chengxu, code set or instruction set, at least one instruction, an at least Duan Chengxu, the code set or
The method that instruction set is loaded by processor and executed to realize the assessment diagnosis of security downside risks.
Therefore the present invention is solved existing by providing a kind of method and system of security downside risks assessment diagnosis
There is a possibility that in technology cannot by scoring obtain to security downside risks future can the loss of energy degree sum understanding, cannot
More accurately comprehensive judgement is made in integrated multidimensional degree information and the similar card Securities price data variation of reference, and the present invention can root
According to the current characteristic information of security, code of points model is automatically generated, is given a mark with the downside risks that code of points is security,
Target security can be put under the visual angle of full market securities according to marking and carry out assessment diagnosis, and further by assessment diagnosis knot
Fruit is converted to portfolio future possible loss;Automatic scoring engine can integrate various dimensions information, while also achieve dynamic
State feature selecting, the function of automatically generating Rating Model and automatic scoring will be improved to security downside risks and portfolio future
Can the loss of energy carry out assessment accuracy.
Detailed description of the invention
To describe the technical solutions in the embodiments of the present invention more clearly, make required in being described below to embodiment
Attached drawing is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the invention, for
For those of ordinary skill in the art, without creative efforts, it can also be obtained according to these attached drawings other
Attached drawing.
Fig. 1 provides a kind of method flow diagram of security downside risks assessment diagnosis for the embodiment of the present invention;
Fig. 2 is the specific flow chart that the embodiment of the present invention is provided as S100;
Fig. 3 is the specific flow chart that the embodiment of the present invention is provided as S200;
Fig. 4 is the specific flow chart that the embodiment of the present invention is provided as S300;
Fig. 5 is the structural schematic diagram of computer provided in an embodiment of the present invention.
Specific embodiment
Below in conjunction with attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that institute
The embodiment of description is only a part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention,
Every other embodiment obtained by those of ordinary skill in the art without making creative efforts, belongs to this hair
The range of bright protection.
In order to solve it is existing in the prior art can not according to existing securities data to it is following can the loss of energy degree and
The understanding of possibility and more accurately comprehensive judgement is made, the present invention provides a kind of assessment diagnosis of security downside risks
Method and system.
Fig. 1 show a kind of flow chart of the method for security downside risks assessment diagnosis of the embodiment of the present invention, such as Fig. 1 institute
It states, the present invention provides a kind of methods of security downside risks assessment diagnosis, comprising:
S100 carries out screening pretreatment to target variable and quantitative characteristic variable in advance, constructs data set.
This step detailed process, as shown in Fig. 2, step S101, describes according to the feature of acquisition, by quantitatively portraying, obtain
To the quantitative characteristic variable;Step S102 is withdrawn according to maximum with descending ranking, obtains the target variable;Step S103,
The data set is constructed according to the time sequencing of the quantitative characteristic variable and the target variable;Step S104, according to structure
The data set built is divided into training set and test set.
In the present embodiment, the quantitative characteristic variable can pass through brainstorming, Zhuan Jiabai by taking public offering fund as an example
Visit, the mode of data to arrange obtains a series of features descriptions about every public offering fund, for example fund manager's fund performance is steady
It is qualitative.Further by the process quantitatively portrayed, these qualitative features are changed into can be with the feature of quantificational expression, as fund is multiple
Weigh the coefficient of variation etc. of net value ranking.The target variable, which refers to a security one day, can calculate one group of quantitative characteristic, than
It is such as one day specified, it uses the maximum of following a period of time security to withdraw based on descending ranking and constructs target variable, work as row
Entitled preceding a quarter is defined as positive sample, is labeled as 1;A quarter is negative sample afterwards, is labeled as 0.Further construct data
Collection, to a security, is specified historical one day, can calculate corresponding quantitative characteristic variate-value and target variable value, and two
Person, which is combined, constitutes a sample;Specifically, the data set built is divided into two parts: training set and test set, tool
Body is to take the sample data of data set the previous year as training set;Take the sample data of data set the latter moon as test set.
It should be noted that above-mentioned following a period of time can be three months or four months, it is above-mentioned that data set the previous year is taken to can be
On December 31,1 day to 2016 January in 2016, above-mentioned take can be 1 day to 2017 4 April in 2017 data set the latter moon
The moon 30, such period can be defined according to actual needs, and this is not limited by the present invention.
S200 constructs automatic scoring engine using the data set, selects dynamic notable feature, establishes dynamic marking mould
Type generates dynamic grading rule, show that final mask scores.
Specifically, step S200 detailed process is as shown in Figure 3:
Step S201, chooses the quantitative characteristic variable in the training set, and input data training pattern is calculated
Weight evidence weight values and information magnitude;Step S202 carries out branch mailbox according to the value of the quantitative characteristic variable and to described quantitative
Characteristic variable screen, output model variable;Step S203 will according to the calculated related coefficient of the model variable
The model variable further screens, and exports final mask variable, obtains having the associated institute of conspicuousness to the target variable
State final mask variable;Step S204 constructs regression model, is intended according to the model variable in conjunction with the target variable
Close equation;Step S205 is fitted the training set according to the fit equation, obtains branch mailbox property parameters and intercept,
By adjusting power formula to update variable's attribute coefficient, the variable's attribute coefficient is in the regression model according to the data set
Variation exports the dynamic grading rule;Step S206, it is original by security according to the branch mailbox property parameters and the intercept
Divide formula to calculate, exports the final mask scoring.
Step S201, chooses the quantitative characteristic variable in the training set, and input data training pattern is calculated
Weight evidence weight values and information magnitude.
In the present embodiment, weight evidence weight values calculation formula are as follows:
Wherein i indicates that independent variable takes i-th of value, and 1 indicates positive sample, 0 table negative sample, y1And y0Respectively indicate target variable
Value is 1 and 0, and n (i) and n (all) respectively indicate record number when taking characteristic variable to take i-th of value and taking all values;Odds withMeaning it is consistent, i.e., fine or not ratio, also referred to as odds.The calculation formula of information magnitude is
Step S202, according to the value of quantitative characteristic variable progress branch mailbox and the progress to the quantitative characteristic variable
Screening, output model variable.
In the present embodiment, according to weight evidence weight values and information magnitude calculation method, to the value of quantitative characteristic variable into
Row branch mailbox simultaneously screens quantitative characteristic.It is specific as follows:
Step 2021, a characteristic variable is chosen from multiple characteristic variables carry out branch mailbox operation;Obtain candidate feature change
Amount.
Step 2022, it takes a candidate feature variable to carry out branch mailbox operation from multiple candidate feature variables, calculates phase
The weight evidence weight values and information magnitude answered determine a preferably branch mailbox mode;In the present embodiment, it can be soundd out according to actual combat gains in depth of comprehension
It chooses, 20,15,10 and 5 demarcation intervals is taken to carry out branch mailbox operation respectively.
Step 2023, judge whether information magnitude is greater than 0.1, if being not more than 0.1, go to step 2021;If it is greater than
0.1, go to step 2024;
Step 2024, judge whether weight evidence weight values are incremented by successively or successively decrease, if not incremented by successively or successively decrease, go to
Step 2025;If it is incremented by successively or successively decrease, step 2026 is gone to;
Step 2025, adjustment branch mailbox section is soundd out, can see make each branch mailbox weight evidence weight values meet order relation, if each point
Case weight evidence weight values are unsatisfactory for order relation, go to step 2021;If each branch mailbox weight evidence weight values meet order relation, step is gone to
2026;
Step 2026, addition individual features variable is further determined whether as model variable there are also candidate feature variable,
If there is no candidate feature variable, terminate;If going to step 2021 there are also candidate feature variable;To which output model becomes
Amount.
Step S203 further screens the model variable according to the calculated related coefficient of the model variable, defeated
Final mask variable out obtains having the associated final mask variable of conspicuousness to the target variable.
In the present embodiment, step S203 specifically comprises the following steps:
Step 2031, initialization model variables set S and final mask variables set S*: the mould screened in step S202
Type variable is put into initialization model variables set S;Initialize final mask variables set S*For sky.
Step 2032, candidate variables are chosen: if S is sky, going to step 2034;It is no, then choose the maximum spy of IV value in S
Variable f is levied, is updated S (S ← S- { f });
Step 2033, it determines final mask variable: if S is sky, f being put into final mask variables set S*In, it goes to step
2034;It is no, then the model variable collection S with f related coefficient more than or equal to 0.6 is selected from S1, in set { f } ∪ S1In select IV value
Maximum characteristic variable f*It is put into final mask variables set S*In, update S (S ← S-S1), go to step 2032;
Step 2034, S is determined*In feature be final mask variable, terminate.
Further, it can dynamically calculating current logarithmic according to the target variable of concentration has the associated final cast of conspicuousness
Variable.
Step S204 constructs regression model according to the model variable in conjunction with the target variable, obtains fit equation;
In the present embodiment, the model variable obtained according to previous step in conjunction with the target variable, return by building
Returning model is logistic regression model, obtains fit equation, specifically, setting y indicates target variable, is constructed and is walked by target variable
The available fit equation of rapid and logistic modular concept is as follows,
Wherein p is the probability of a record preferably, xijIt indicates j-th of branch mailbox attribute value of ith feature variable (taking 0 or 1),
ωijIndicate the parameter of j-th of branch mailbox attribute of ith feature variable, b indicates intercept, and m indicates Characteristic Number.With in previous step
Fit equation training dataset is fitted, can be in the hope ofAnd b*。
Step S205 is fitted the training set according to the fit equation, obtains branch mailbox property parameters and intercept,
By adjusting power formula to update variable's attribute coefficient, the variable's attribute coefficient is in the logistic regression model according to
The variation of data set exports the dynamic grading rule.
In the present embodiment, formula is weighed by adjustingUpdate each variable's attribute
Coefficient.These variable's attribute coefficients are different with the difference of data set, constitute the marking rule of one group of dynamic adjustment.
Further, it is obtained according to preceding stepAnd b*, security X can be calculated by following formulaiOriginal point:
The final mask scoring that benchmark is divided into 200 is converted to original point by following formula again.
Wherein, for quality than being 1:1,20 be double odds incremental value.
The final mask is scored and inputs autotext diagnostor by S300, calculates standard security point, and it is literary to generate diagnosis
This.
This step detailed process, as shown in figure 4, step S301, obtains low-risk probability P according to scorecard odds principle
Value;Specifically, according to scorecard principle, automatic scoring engine marking s=200+20x, so as to calculateAccording to
It can be calculated according to scorecard odds principle
The low-risk probability P value is distinguished input linear interpolating estimation method and partition balancing estimation side by step S302
The two different evaluation methods of method, method particularly includes:
I. linear interpolation evaluation method
Ii. partition balancing evaluation method
Output Pi (A) value and Pi (B) value respectively choose the average value of the Pi (A) value and Pi (B) value as final probability
DistributionObtain risk ranking probability distribution P (x);
Step S303 obtains target security maximum in different time periods and withdraws rate, divides calculation formula according to standard security, obtain
To the standard security point of the target security, specific formula are as follows:
Wherein x is the MDD of different level, and p (x) is corresponding probability distribution.
Step S304, according to the score value of the dynamic grading Rule security items distinguished variable, by institute's commentary
Score value compares, and is equipped with evaluation text to every distinguished variable in advance, finds correspondence according to the standard security point
Evaluation text obtain assessment diagnostic text.
In an implementation, according to newest training data dynamic generation scoring model, and then marking rule is obtained.According to marking
Rule can obtain marking of each security on each distinguished variable, and lateral comparison is done in these marking, can determine whether out to be evaluated
Security are estimated in relatively fine or not degree entirely in the market.Fine or not two classes evaluation can be pre-designed to each distinguished variable, as long as
Given a mark accordingly according to security to be evaluated find it is corresponding quality evaluation can obtain security to be evaluated in each notable feature
Tentative diagnosis text.Further, it is possible to be handled by polishing appropriate, so that it may generate corresponding assessment diagnostic text.
In specific embodiment, by taking public offering fund as an example, the fund that on January 14th, 2019 is 001071 to code is integrated
Diagnosis generates diagnostic text." marking of fund standard security are as follows: 90.7 points, high risk.Fund three months possible maximums of future
Withdraw is 7%.The whole below average probability of the ability to ward off risks are as follows: 82.6%.' nearly 1 Nian Fuquan net value increase ' score value is very
Height, past amount of increase and amount of decrease are more moderate.' investigate that nearly value ' the score value of style variation in 1 month is too low, and recent style performance is unstable year,
Fund risk is bad to be estimated.' fund manager holds a post, and time ' score value is too low, and fund manager's tenure factor will generate risk certain
Negative effect.' closely March, stability bandwidth ' score value was too low, recent amount of increase and amount of decrease fluctuation is excessively high, and risk is larger.' fund classification ' score value mistake
Low, generic overall risk under Vehicles Collected from Market market is higher.
In conclusion the present invention can be current according to security characteristic information, code of points model is automatically generated, with scoring
Rule is that the downside risks of security are given a mark, and can be put into target security under the visual angle of full market securities and carry out according to marking
Assessment diagnosis, and assessment diagnostic result is further converted into portfolio future possible loss.Automatic scoring engine can
The function of integrating various dimensions information, while also achieving dynamic feature selection, automatically generating Rating Model and automatic scoring, and it is existing
There is technology to compare, credit scoring technology is applied in the prediction of security downside risks, improves and security downside risks and security is provided
Produce it is following can the assessment accuracy that carries out of the loss of energy.
Based on the same technical idea, it is the embodiment of the invention also provides a kind of assessment diagnosis of security downside risks
System, the method for executing above-mentioned security downside risks assessment diagnosis.
It should be understood that the system of security downside risks assessment diagnosis provided by the above embodiment, in practical applications,
It can according to need and be completed by different functional modules above-mentioned function distribution, i.e., the internal structure of system is divided into difference
Functional module, to complete all or part of the functions described above.In addition, security downside risks provided by the above embodiment
The embodiment of the method for the system and the assessment diagnosis of security downside risks of assessing diagnosis belongs to same design, and specific implementation process is detailed
See embodiment of the method, which is not described herein again.
Fig. 5 is the structural schematic diagram of computer provided in an embodiment of the present invention.The computer 500 can because configuration or performance not
With and generate bigger difference, may include one or more central processing units 522 (for example, one or more
Processor) and memory 532, the storage medium 530 (such as one of one or more storage application programs 542 or data 544
A or more than one mass memory unit).Wherein, memory 532 and storage medium 530 can be of short duration storage or persistently deposit
Storage.The program for being stored in storage medium 530 may include one or more modules (diagram does not mark), and each module can be with
Including being operated to the series of instructions in computer 500.Further, central processing unit 522 can be set to be situated between with storage
Matter 530 communicates, and the series of instructions operation in storage medium 530 is executed on computer 500.
Computer 500 can also include one or more power supplys 526, one or more wired or wireless networks
Interface 550, one or more input/output interfaces 558, one or more keyboards 556, and/or, one or one
The above operating system 541, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD etc..
Computer 500 may include having perhaps one of them or one of more than one program of memory and one
Procedure above is stored in memory, and is configured to execute one or one by one or more than one processor
Procedure above includes the instruction for carrying out above-mentioned security downside risks assessment diagnosis.
Those of ordinary skill in the art will appreciate that realizing that all or part of the steps of above-described embodiment can pass through hardware
It completes, relevant hardware can also be instructed to complete by program, the program can store in a kind of computer-readable
In storage medium, storage medium mentioned above can be read-only memory, disk or CD etc..
Through the above description of the embodiments, those skilled in the art can be understood that each embodiment can
It realizes by means of software and necessary general hardware platform, naturally it is also possible to pass through hardware.Based on this understanding, on
Stating technical solution, substantially the part that contributes to existing technology can be embodied in the form of software products in other words, should
Computer software product may be stored in a computer readable storage medium, such as ROM/RAM, magnetic disk, CD, including several fingers
It enables and using so that a computer equipment (can be personal computer, server-side or the network equipment etc.) executes each implementation
Method described in certain parts of example or embodiment.
The foregoing is merely presently preferred embodiments of the present invention, is not intended to limit the invention, it is all in spirit of the invention and
Within principle, any modification, equivalent replacement, improvement and so on be should all be included in the protection scope of the present invention.
Claims (10)
1. a kind of method of security downside risks assessment diagnosis characterized by comprising
Data set is constructed according to target variable and quantitative characteristic variable;
According to the collection selection dynamic notable feature, model variable is obtained, dynamic is established according to the model variable and is given a mark
Model generates code of points according to the dynamic scoring model, obtains final mask score value according to the code of points;
Low-risk probability value and final probability distribution value are calculated according to the final mask score value, obtains standard security point, root
According to the standard security point, diagnostic text is obtained.
2. the method for security downside risks assessment diagnosis according to claim 1, which is characterized in that described to be become according to target
The step of amount constructs data set with quantitative characteristic variable specifically includes:
Feature description is obtained, and feature description is quantitatively portrayed, obtains the quantitative characteristic variable;
It is withdrawn according to maximum with descending ranking, obtains the target variable;
The data set is constructed according to the time sequencing of the quantitative characteristic variable and the target variable;
The data set is divided into training set and test set according to time sequencing standard.
3. the method according to claim 1, wherein described according to the collection selection dynamic notable feature,
Model variable is obtained, dynamic scoring model is established according to the model variable, scoring rule are generated according to the dynamic scoring model
Then, the step of obtaining final mask score value according to the code of points specifically includes:
Choose the quantitative characteristic variable in the training set, input data training pattern, be calculated weight evidence weight values with
Information magnitude;
Branch mailbox and screening to the quantitative characteristic variable, output model are carried out according to the value of the quantitative characteristic variable
Variable.
4. according to the method described in claim 3, it is characterized in that, described according to the collection selection dynamic notable feature,
Model variable is obtained, dynamic scoring model is established according to the model variable, scoring rule are generated according to the dynamic scoring model
Then, the step of final mask score value being obtained according to the code of points further include:
According to the calculated related coefficient of the model variable, the model variable is further screened, output final mask becomes
Amount obtains having the associated final mask variable of conspicuousness to the target variable;
According to the model variable in conjunction with the target variable, regression model is constructed, obtains fit equation.
5. according to the method described in claim 4, it is characterized in that, described according to the collection selection dynamic notable feature,
Model variable is obtained, dynamic scoring model is established according to the model variable, scoring rule are generated according to the dynamic scoring model
Then, the step of final mask score value being obtained according to the code of points further include:
The training set is fitted according to the fit equation, obtains branch mailbox property parameters and intercept, weighs formula by adjusting
Variable's attribute coefficient is updated, the variable's attribute coefficient is in the regression model according to the variation of data set output
Dynamic grading rule;
According to the branch mailbox property parameters and the intercept, is calculated by original point of formula of security, export the final mask and comment
Point.
6. according to the method described in claim 5, it is characterized in that, described calculate low-risk according to the final mask score value
The step of probability value and final probability distribution value obtain standard security point, divided according to the standard security, obtain diagnostic text tool
Body includes:
Low-risk probability value is obtained according to scorecard odds principle;
By low-risk probability value input linear interpolating estimation method and partition balancing evaluation method is two different estimates respectively
Calculation method exports the low-risk probability value of two kinds of evaluation methods respectively, takes out average value, obtains risk ranking probability distribution;
It obtains target security maximum in different time periods and withdraws rate, divide calculation formula according to standard security, obtain the target card
The standard security of certificate point;
According to the score value of the dynamic grading Rule security items distinguished variable, the score value is compared,
Evaluation text is equipped with to every distinguished variable in advance, corresponding evaluation text is found according to the standard security point and is obtained
Assess diagnostic text.
7. according to the method described in claim 4, output is most it is characterized in that, described further screen the model variable
Final cast variable, the step of obtaining having the target variable conspicuousness associated final mask variable, specifically include:
The model variable input initialization model variable is concentrated, the final mask variables set of output;
Choose the candidate model variable;
Determine the final mask variable;
Determine that the model variable in the final mask variables set is the final mask variable.
8. a kind of system of security downside risks assessment diagnosis, which is characterized in that for executing such as any one of claim 1 to 7
The method of the security downside risks assessment diagnosis.
9. a kind of computer, which is characterized in that the computer includes processor and memory, be stored in the memory to
Few an instruction, at least a Duan Chengxu, code set or instruction set, it is at least one instruction, an at least Duan Chengxu, described
Code set or instruction set are loaded by the processor and are executed to realize security downlink as described in any one of claim 1 to 7
The method of risk assessment diagnosis.
10. a kind of computer readable storage medium, which is characterized in that be stored at least one instruction, extremely in the storage medium
A few Duan Chengxu, code set or instruction set, at least one instruction, an at least Duan Chengxu, the code set or instruction
Collection is loaded by processor and is executed the side to realize security downside risks assessment diagnosis as described in any one of claim 1 to 7
Method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910277134.0A CN109978406A (en) | 2019-04-08 | 2019-04-08 | A kind of method and system of security downside risks assessment diagnosis |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910277134.0A CN109978406A (en) | 2019-04-08 | 2019-04-08 | A kind of method and system of security downside risks assessment diagnosis |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109978406A true CN109978406A (en) | 2019-07-05 |
Family
ID=67083419
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910277134.0A Pending CN109978406A (en) | 2019-04-08 | 2019-04-08 | A kind of method and system of security downside risks assessment diagnosis |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109978406A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112801563A (en) * | 2021-04-14 | 2021-05-14 | 支付宝(杭州)信息技术有限公司 | Risk assessment method and device |
CN113205880A (en) * | 2021-04-30 | 2021-08-03 | 广东省人民医院 | LogitBoost-based heart disease prognosis prediction method and device |
-
2019
- 2019-04-08 CN CN201910277134.0A patent/CN109978406A/en active Pending
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112801563A (en) * | 2021-04-14 | 2021-05-14 | 支付宝(杭州)信息技术有限公司 | Risk assessment method and device |
CN113205880A (en) * | 2021-04-30 | 2021-08-03 | 广东省人民医院 | LogitBoost-based heart disease prognosis prediction method and device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110298663A (en) | Based on the wide fraudulent trading detection method learnt deeply of sequence | |
CN106156809A (en) | For updating the method and device of disaggregated model | |
Dash | An improved shuffled frog leaping algorithm based evolutionary framework for currency exchange rate prediction | |
CN107545038A (en) | A kind of file classification method and equipment | |
CN109978406A (en) | A kind of method and system of security downside risks assessment diagnosis | |
CN111062806A (en) | Personal finance credit risk evaluation method, system and storage medium | |
Sampath et al. | A generalized decision support framework for large‐scale project portfolio decisions | |
Kannan et al. | A novel software package selection method using teaching–learning based optimization and multiple criteria decision making | |
CN110517142B (en) | Method and device for outputting policy evaluation information | |
Voynarenko et al. | Applying fuzzy logic to modeling economic emergence | |
Srivastava et al. | Efficiency evaluation of assets and optimal portfolio generation by cross efficiency and cumulative prospect theory | |
CN109919374A (en) | Prediction of Stock Price method based on APSO-BP neural network | |
Sheel | Monte Carlo Simulations and Sceenrio Analyst— Decision-Making Tools for Hoteliers | |
Shumway | Allocation of scarce resources to agricultural research: Review of methodology | |
CN107274043A (en) | Quality evaluating method, device and the electronic equipment of forecast model | |
CN110322342A (en) | Borrow or lend money construction method, system and the debt-credit Risk Forecast Method of risk forecast model | |
Ballin et al. | Optimization of sampling strata with the SamplingStrata package | |
CN107169571A (en) | A kind of Feature Selection method and device | |
Wolfe | How To Adjust Forecast With The Analytic Hierarachy Process | |
CN106776892A (en) | Based on music platform data assessment musical works network attention data method and system | |
KR101099873B1 (en) | Stratification Benchmarking System And method thereof | |
CN110443374A (en) | A kind of resource information processing method, device and equipment | |
CN110991600A (en) | Drought intelligent prediction method integrating distribution estimation algorithm and extreme learning machine | |
Podkopaev et al. | An approach to the automatic comparison of reference point-based interactive methods for multiobjective optimization | |
Truscott et al. | Detecting shadow economy sizes with symbolic regression |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190705 |
|
RJ01 | Rejection of invention patent application after publication |