CN106067094A - A kind of dynamic assessment method and system - Google Patents
A kind of dynamic assessment method and system Download PDFInfo
- Publication number
- CN106067094A CN106067094A CN201610416226.9A CN201610416226A CN106067094A CN 106067094 A CN106067094 A CN 106067094A CN 201610416226 A CN201610416226 A CN 201610416226A CN 106067094 A CN106067094 A CN 106067094A
- Authority
- CN
- China
- Prior art keywords
- model
- evaluation
- assessment
- evaluation object
- 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
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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/28—Databases characterised by their database models, e.g. relational or object models
- G06F16/283—Multi-dimensional databases or data warehouses, e.g. MOLAP or ROLAP
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Bioinformatics & Computational Biology (AREA)
- Entrepreneurship & Innovation (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Evolutionary Computation (AREA)
- Educational Administration (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Probability & Statistics with Applications (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of dynamic assessment method and system, the method includes: pre-build the dynamic evaluation model system based on intelligence decision support system system, and dynamic evaluation model system includes: based on the man-machine interactive system of the open dynamic data base system of teledata search pattern, disaggregated model storehouse system, intelligent evaluation system and the visualization appraisal framework based on single interface encapsulation.Intelligent evaluation system is used for calling, managing dynamic data base system and disaggregated model storehouse system, and to carry out information mutual with man-machine interactive system.Intelligent evaluation system is by judging the classification of evaluation object and select assessment models and related assessment parameter accordingly, and chooses optimal models in assessment models storehouse and carry out equity assessment to company to be assessed.The Evaluation Platform of a circulation can be built by implementing the present invention, in conjunction with having simple and convenient relative technique of estimation and objective and accurate absolute technique of estimation, form the Intelligent Dynamic evaluation system of complete set.
Description
Technical field
The present invention relates to computer intelligence assessment technology field, particularly to a kind of dynamic assessment method and dynamic evaluation mould
Type system.
Background technology
In recent years, intelligence decision support system system (IDSS, Intelligent Decision Support System) oneself becomes
System engineering and the important subject in computer application field.By numerous associated specialist, constantly the exploring and grind of scholar
Studying carefully, IDSS has obtained fast development in the theoretical research of academia and the actual application of national economy.Since IDSS exploitation,
Compared with its theoretical research, actual application work is carried out earlier, and it is widely used in business administration, system development, economy point
The aspects such as analysis and planning, strategic research, resource management, investment planning, support that the DSS of all kinds of decision problem is a large amount of
Occur and come into operation.
From the point of view of Gai Kuoing, IDSS is based on management science, operational research, cybernetics and behavior science, with computer technology,
Artificial intelligence technology and information technology are means, intelligently support the computer system of decision making.It passes through human-computer dialogue
It is analyzed, compares and judges, and then identify problem, set up or modification model, the clear and definite decision objective of aid decision making person, be decision-making
Person provides various scheme and carries out to it evaluating and preferably, provide beneficial help for correct decisions.
But, intelligent financial system is confined to utilize chart to be described current condition also mostly, and solving, enterprise is general
Time management and decision problem aspect be still within the research and probe stage.Here, as a example by enterprise's value of stock right assessment, current
Value of stock right assessment is based primarily upon the assessment analysis to a large amount of financial datas for the participant, to assess knowledge, the warp of participant self
Test, carry out very subjective assessment based on the factor such as technology to the value of stock right, do so not only time-consuming but also effort, and also accurate not
Really, intelligence.
Content of the invention
In view of this, the purpose of the embodiment of the present invention is to propose a kind of dynamic assessment method and dynamic evaluation model system
System, designs the evaluation system of a kind of Intelligent Dynamic, is capable of efficiently assessing accurately.
From the point of view of further, this dynamic assessment method comprises the following steps: pre-build based on intelligence decision support system system
(IDSS) dynamic evaluation model system, described dynamic evaluation model system includes: based on the opening of teledata search pattern
Formula dynamic data base system, disaggregated model storehouse system, intelligent evaluation system and the visualization assessment frame based on single interface encapsulation
The man-machine interactive system of frame;Described intelligent evaluation system obtains the identity information of evaluation object by man-machine interactive system;Call
Described dynamic data base system, described dynamic data base system is according to the identity information of described evaluation object, described intelligent evaluation
Screening conditions, searching route and search keyword search that system is arranged simultaneously obtain target data institute with regard to described evaluation object
Html source code, and Data Analysis and reconstruct are carried out to described html source code;
Calling described disaggregated model storehouse system, described disaggregated model storehouse system obtains according to described dynamic data base system
Target data, extracts and can reflect the structural characterization data of described evaluation object, described structural characterization data is identified and
Classification analysis computing;Described intelligent evaluation system is according to the operation result of described disaggregated model storehouse system, it is judged that described assessment is right
As if belong to outliers and still fall within gregarious sample, and choose corresponding optimum evaluation model described evaluation object is commented
Estimate;If described evaluation object belongs to gregarious sample, then the assessment parameter according to described classification analysis result and setting, at described point
Class model storehouse system calls the assessment models matching with described evaluation object, uses relative valuation mode, to described assessment
Object is estimated, and exports assessment result;If described evaluation object belongs to outliers, then according to described classification analysis result
And the assessment parameter arranging, use absolute valuation mode, described evaluation object is estimated, and exports assessment result.
Alternatively, in certain embodiments, described dynamic data base system is according to the identity information of described evaluation object, institute
State screening conditions, searching route and the search keyword search of intelligent evaluation system setting and obtain with regard to described evaluation object
The html source code at target data place, and Data Analysis is carried out to described html source code and reconstruct includes: right according to described assessment
The identity information of elephant selectes evaluation object, obtains corresponding screening conditions, searching route and the search keyword pre-setting;Institute
State dynamic data base system and use the remote online Database pattern based on open crawler algorithm, when described evaluation object
After Xuan Ding, assess requisite number by open crawler algorithm according to evaluation object relevant information and default URL site search
According to, by keyword response in the way of access one or more remote server port, collect and download corresponding target data institute
Html source code;Wherein, described remote server port is stored in described dynamic data base system, is called by URL
Mode accesses target pages;Resolve by way of regular expression character string tagsort, download described html source code to this
Ground server, thus the dynamic interim remote online database of form one;Wherein, described regular expression and described number of targets
According to keyword match.
Alternatively, in certain embodiments, described disaggregated model storehouse system is according to described dynamic data base system, extracts energy
Enough reflect the structural characterization data of described evaluation object, described structural characterization data is identified and classification analysis includes: adopt
Extract with the structural characterization data to described evaluation object for the Data Analysis Model pre-building;Wherein, described data are divided
Analysis model includes for principal component analysis (PCA) model carrying out Feature Compression and dimension reduction to described evaluation object;Adopt
With the mixed model pre-building, classification analysis is carried out to described evaluation object;Wherein, described mixed model includes: hierarchical clustering
Model and SVMs (SVM) model, described hierarchical clustering model is used for identifying described evaluation object and similar references object
Correlation degree and stamp the hierarchical clustering model of correlation tag for this correlation degree;SVMs (SVM) model is used for
According to described correlation tag judge described evaluation object whether can match can as the similar references object of assessment reference, if
Described evaluation object can match can be as the similar references object of assessment reference, then described evaluation object belongs to gregarious sample,
Otherwise described evaluation object belongs to outliers.
Alternatively, in certain embodiments, described disaggregated model storehouse is for providing it to store for described intelligent evaluation system
Each class model, described disaggregated model storehouse is also configured with model base management system, and described model base management system is for storage
Model carry out model extraction, access, update and synthetic operation;Wherein, described disaggregated model storehouse sets for similar references object
It is equipped with special purpose model, routine analysis model, built-up pattern and temporary pattern;Described disaggregated model storehouse is also configured with model depositary management
Reason system, described model base management system includes construction management module, access management module, operational management module, is additionally provided with
Model dictionary, internal database and external data base, for the correlation model managing, calling, configure in described disaggregated model storehouse.
Alternatively, in certain embodiments, described evaluation object is the value of stock right of Target Enterprise, described evaluation object
Identity information is CompanyName or stock code;Described relative valuation mode includes p/e ratio (P/E) model, HSBC (P/B)
Model and city's pin rate (P/S) model;Described intelligent evaluation system is according to the assessment ginseng designing based on knowwhy and expertise
Number, chooses corresponding optimum evaluation model and carries out share price value assessment to the evaluation object belonging to gregarious sample;Described definitely estimate
Value mode includes that free cash flow is discounted (NPV) model, option on physicals (ROpt.) Valuation Modelling;Described intelligent evaluation system according to
The assessment parameter designing based on knowwhy and expertise, chooses corresponding optimum evaluation model and comments to belonging to outliers
Estimate object and carry out share price value assessment;Wherein, described hierarchical clustering model is for for analyzing the Kmeans of the relevant enterprise degree of association
Clustering Model.
Alternatively, in certain embodiments, the screening conditions that described intelligent evaluation system is arranged for based on knowwhy and
The reasoning from logic principle that expertise designs, is packaged in described intelligent evaluation system with code form, and carries out in order
Coding, can be normalized, systematized call.
For realizing above-mentioned dynamic assessment method, the present invention also proposes a kind of dynamic evaluation model system.From the point of view of further, should
Dynamic evaluation model system includes: man-machine interactive system, based on single interface encapsulation, has visualization appraisal framework, is provided with people
Machine interactive interface and result output display interface;Dynamic data base system, based on teledata search pattern, has open dynamic
State online database, closes for the identity information according to described evaluation object, set screening conditions, searching route and search
The html source code at the target data place with regard to described evaluation object is searched for and obtained to keyword, and enters line number to described html source code
According to parsing and reconstruct;Disaggregated model storehouse system, for the target data obtaining according to described dynamic data base system, extraction can
Reflect the structural characterization data of described evaluation object, described structural characterization data is identified and classification analysis computing;Intelligence
Assessment system, for the operation result according to described disaggregated model storehouse system, it is judged that described evaluation object is belonging to outliers
Still fall within gregarious sample, and choose corresponding optimum evaluation model described evaluation object is estimated;When described assessment is right
During as belonging to gregarious sample, for the assessment parameter according to described classification analysis result and setting, in described disaggregated model storehouse system
System calls the assessment models matching with described evaluation object, uses relative valuation mode, described evaluation object is commented
Estimate, and export assessment result;And, when described evaluation object belongs to outliers, for according to described classification analysis result
And the assessment parameter arranging, use absolute valuation mode, described evaluation object is estimated, and exports assessment result.
Alternatively, in certain embodiments, described dynamic data base system includes: open remote online database,
Local data base and data base management system;Described dynamic data base system uses open crawler algorithm as building storehouse engine;
And/or, described disaggregated model storehouse system includes disaggregated model bunch and model base management system, and described disaggregated model bunch is described
Intelligent evaluation system provides its each class model storing, and described model base management system carries for carrying out model to the model of storage
Take, access, update and synthetic operation;Described disaggregated model storehouse system is for using the Data Analysis Model that pre-builds to described
The structural characterization data of evaluation object carries out extracting and for using the mixed model pre-building to enter described evaluation object
Row classification analysis;Wherein, described disaggregated model bunch includes Data Analysis Model, and described Data Analysis Model includes for described
Principal component analysis (PCA) model carrying out Feature Compression and dimension reduction of evaluation object;Described mixed model includes: level gathers
Class model and SVMs (SVM) model, described hierarchical clustering model is used for identifying described evaluation object and similar reference pair
The correlation degree of elephant the hierarchical clustering model stamping correlation tag for this correlation degree;SVMs (SVM) model is used
According to described correlation tag judge described evaluation object whether can match can as the similar references object of assessment reference, as
Really described evaluation object can match can be as the similar references object of assessment reference, then described evaluation object belongs to gregarious sample
This, otherwise described evaluation object belongs to outliers;Described disaggregated model bunch also includes value of stock right assessment models, described equity
Appraisal Model includes: p/e ratio (P/E) model in the first model library, HSBC (P/B) model, city's pin rate (P/S) mould
Type;And, the free cash flow in the second model library is discounted (NPV) model and option on physicals (ROpt.);Wherein, described first
The assessment parameter that in model library and described second model library, the Call Condition of each model is arranged by described intelligent evaluation system determines.
Alternatively, in certain embodiments, described intelligent evaluation system is used for calling, manages described dynamic data base system
And disaggregated model storehouse system, and to carry out information mutual with described man-machine interactive system;Described intelligent evaluation system includes: body is led
Territory module, for providing the shared ideas model to correlation theory and Formal Specification explanation;Reasoning from logic module: be used for providing
Correlation theory and the matching way of corresponding logical relation;Task calculating module: based on carrying out according to the Valuation Modelling determining
Calculate and result of calculation is exported.
Alternatively, in certain embodiments, the screening conditions that described intelligent evaluation system is arranged for based on knowwhy and
The reasoning from logic principle that expertise designs, is packaged in described intelligent evaluation system with code form, and carries out in order
Coding, can be normalized, systematized call.
Relative to prior art, various embodiments of the present invention have the advantage that
After using the technical scheme of the embodiment of the present invention, by introducing information search, data mining, expertise reasoning etc.
Technology, is worth for listed company equities and sets up a set of dynamic evaluation model system based on IDSS, this dynamic evaluation model system
System includes dynamic data base system, disaggregated model storehouse system, intelligent evaluation system and man-machine interactive system.Wherein, intelligence is commented
Estimate system for calling, manage described dynamic data base system and disaggregated model storehouse system, and enter with described man-machine interactive system
Row information is mutual.Intelligent evaluation system uses the selection to Valuation Modelling for a kind of condition judgment based on binary tree to make inferences
Screening, judges whether evaluation object also exists higher similarity with reference to assessment models, and chooses in assessment models storehouse
Optimal models carries out equity assessment to company to be assessed.The present invention uses mathematical logic and algorithm to deduce, and builds a circulation
Evaluation Platform, in conjunction with have simplification, facilitation and the features such as ageing relative technique of estimation with there is objectivity, accuracy
And the absolute technique of estimation of feature such as completeness, form the Intelligent Dynamic evaluation system of complete set.And it is open by design one
The remote online Database Systems of formula, when making value assessment, it is not necessary to manually goes to collect the data required for assessment, and passes through
The unstructured data crawling on networking is resolved to structural data by regular expression, under the structural data that finally will obtain
Being downloaded to a volatile data base of this locality, the data as assessment are supported.So both evaded produced by traditional Manpower Estimate
Subjective error, reduces again the cost loss that evaluation process is shone, to greatest extent by manpower from heavy evaluation work
Free.
More feature of the embodiment of the present invention and advantage will be explained in detailed description of the invention afterwards.
Brief description
Constitute a part of accompanying drawing of the embodiment of the present invention to be used for providing being further appreciated by the embodiment of the present invention, the present invention
Schematic description and description be used for explaining the present invention, be not intended that inappropriate limitation of the present invention.In the accompanying drawings:
The schematic flow sheet of the dynamic assessment method that Fig. 1 provides for the embodiment of the present invention;
The composition frame chart of the dynamic evaluation model system that Fig. 2 provides for the embodiment of the present invention.
Detailed description of the invention
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Describe, it is clear that described embodiment is only a part of embodiment of the present invention, rather than whole embodiments wholely.Based on
Embodiment in the present invention, it is every other that those of ordinary skill in the art are obtained under the premise of not making creative work
Embodiment, broadly falls into the scope of protection of the invention.
It should be noted that in the case of not conflicting, feature in the embodiment of the present invention and embodiment can mutual group
Close.
Below in conjunction with the accompanying drawings, various embodiments of the present invention are described further:
Embodiment of the method
For solving, the embodiment of the present invention proposes a kind of dynamic assessment method, with reference to the schematic flow sheet shown in Fig. 1, the party
Method include following process step:
S102: pre-build the dynamic evaluation model system based on intelligence decision support system system (IDSS), dynamic evaluation model
System includes: based on the open dynamic data base system of teledata search pattern, disaggregated model storehouse system, intelligent evaluation system
System and the man-machine interactive system of the visualization appraisal framework based on single interface encapsulation.
S104: intelligent evaluation system obtains the identity information of evaluation object by man-machine interactive system.
Wherein, man-machine interactive system uses the I/O design architecture of the visualization appraisal framework (SIPVF) of single interface encapsulation,
Can greatly simplify evaluation process, improve the efficiency of assessment.
S106: call dynamic data base system, dynamic data base system is according to the identity information of evaluation object, intelligent evaluation
The screening conditions of system setting, searching route and search keyword search simultaneously obtain the target data place with regard to evaluation object
Html source code, and Data Analysis and reconstruct are carried out to html source code.
In this step, dynamic data base system uses the online database based on open crawler algorithm and builds storehouse mould
Formula, the data warehouse using such online database mode of establishing database to set up is dynamic, OO, i.e. works as evaluation object
After Xuan Ding, the crawler algorithm of model library can according to evaluation object, and be previously set URL site search assessment requisite number
According to, and resolved by regular expression, download these data to home server, thus one dynamic ephemeral data of form
Storehouse.This online database based on open crawler algorithm is advantageous in that: it had both improve the efficiency of assessment, ensures again to comment
Estimate the quality of data, and this storehouse mode of building is applicable to nearly all listed company.
S108: calling classification model-base management system, the number of targets that disaggregated model storehouse system obtains according to dynamic data base system
According to extraction can reflect the structural characterization data of evaluation object, is identified structural characterization data and classification analysis computing;
S110: intelligent evaluation system is according to the operation result of disaggregated model storehouse system, it is judged that evaluation object is belonging to peel off
Sample still falls within gregarious sample, and chooses corresponding optimum evaluation model and be estimated evaluation object, if evaluation object belongs to
In gregarious sample, then step S112;If evaluation object belongs to outliers, then step S114;
S112: the assessment parameter according to classification analysis result and setting, call in the system of disaggregated model storehouse with assess right
As the assessment models matching, use relative valuation mode, evaluation object is estimated, and exports assessment result;
S114: the assessment parameter according to classification analysis result and setting, uses absolute valuation mode, carries out evaluation object
Assessment, and export assessment result.
Above-described embodiment establishes the basic framework of the listed company equities Value accounting system based on IDSS, uses mathematics
Logic and algorithm deduce the Evaluation Platform building circulation, by judging the classification of evaluation object and selecting assessment models and phase accordingly
The assessment parameter closed, and in assessment models storehouse, choose optimal estimation model equity assessment is carried out to company to be assessed, so
Can substantially realize intellectuality, the automation assessment that all kinds of listed company equities are worth, in conjunction with having simplification, facilitation
With the relative technique of estimation of the feature such as ageing with there is the absolute technique of estimation of feature such as objectivity, accuracy and completeness, form one
Overlap complete Intelligent Dynamic evaluation system, this special case filter-type reasoning pattern, not only meet the assessment of ordinary enterprises, moreover it is possible to look for
Go out the equity assessment models that special enterprises is suitable for.Therefore, above-described embodiment had both been evaded produced by traditional Manpower Estimate main
See error, reduce again the cost loss that evaluation process is shone to greatest extent, manpower is solved from heavy evaluation work
Release.
As the optional embodiment of one, in above-described embodiment S106, dynamic data base system is according to evaluation object
Identity information, intelligent evaluation system arrange screening conditions, searching route and search keyword search and obtain with regard to assessment right
The html source code at the target data place of elephant, and carry out Data Analysis to html source code and the step of reconstruct includes following process
Journey:
S1061: select evaluation object according to the identity information of evaluation object, obtains the corresponding screening bar pre-setting
Part, searching route and search keyword;
S1062: dynamic data base system uses the remote online Database pattern based on open crawler algorithm, when
After evaluation object is selected, commented according to evaluation object relevant information and default URL site search by open crawler algorithm
Estimate desired data, in the way of keyword response, access one or more remote server port, collect and download corresponding mesh
The html source code at mark data place;Wherein, remote server port is stored in dynamic data base system, is called by URL
Mode accesses target pages;
S1063: resolve by way of regular expression character string tagsort, download html source code to local service
Device, thus the dynamic interim remote online database of form one;Wherein, the keyword phase of regular expression and target data
Join.
In above-described embodiment, as a example by evaluation object is as listed company, built by using the web crawlers algorithm of intelligence
Open dynamic data base system, this dynamic data base system is by the title according to evaluation object or keyword (such as: listed company
Stock code) and screening conditions etc., by way of a kind of keyword responds, access one (or several) remote server port,
On the internet URL (uniform resource locator) port of evaluation object and industry associated companies Financial Information thereof is searched
Web character string is simultaneously resolved by rope by regular expression, extracts crucial financial data, and locally downloading.Wherein,
Remote server port refers to some finance and economic databases increased income, portal website, the API port of resource center.These API
Port will be stored in a nucleus module of IDSS data warehouse in advance, and comes by way of URL calls
Access.And these finance and economic databases, portal website, the API port of resource center typically all cover substantial amounts of dissimilar
Corporate financial data, in each THML file that they are distributed under web-site, and each other exist association, therefore,
Use crawler algorithm can find it one by one and download.
Below regular expression is once illustrated: in above-described embodiment, by crawler algorithm, we can obtain number of targets
According to the html source code at place, but html source code includes substantial amounts of redundancy.Through observation shows that, the code in HTML
It is that the form with character string shows, and these character strings have specific architectural feature mostly.Source code quilt due to HTML
It is embedded in character string, it is impossible to be directly drawn off and be stored in the database of this locality, therefore, use one to be used for character string
The syntax rule (i.e. regular expression) of coupling, resolves the character string of source code in HTML, in order to extract in character string useful
Data.Here, regular expression also known as normal representation method (Regular Expression, be often abbreviated as in code regex,
Regexp or RE), it is a kind of method of description character trail concentrating the common characteristic of each character string as foundation with character string.Just
Then expression formula may be used for search, editor or operation text and data.Herein, regular expression is listed by coupling
The keyword (as shown in following table) of the financial data needed for stock price value assessment, and concrete according to these keyword searches
Financial data.
Regular expression escape character list
Upper table gives regular expression conventional escape character, and the combination utilizing these escape characters is constituted regular expressions
The grammer of formula, and extract the information required for assessment with this grammer.
As the optional embodiment of one, in above-described embodiment S108, disaggregated model storehouse system is according to dynamic data base
System, extracts and can reflect the structural characterization data of evaluation object, is identified structural characterization data and the step of classification analysis
Suddenly following handling process is included:
S1081: use the structural characterization data to evaluation object for the Data Analysis Model pre-building to extract;
Wherein, Data Analysis Model include for evaluation object carry out Feature Compression and the principal component of dimension reduction is divided
Analysis (Principal Component Analysis, PCA) model.
S1082: use the mixed model pre-building to carry out classification analysis to evaluation object;
Wherein, mixed model includes hierarchical clustering model and SVMs (SVM) model, and hierarchical clustering model is used for knowing
Other evaluation object and the correlation degree of similar references object the hierarchical clustering mould stamping correlation tag for this correlation degree
Type;SVMs (SVM) model can be as assessment reference for judging whether evaluation object can match according to correlation tag
Similar references object, if evaluation object can match can as the similar references object of assessment reference, then evaluation object belong to
In gregarious sample, otherwise evaluation object belongs to outliers.
Below, pca model is described further:
In order to carry out objectively analyzing to evaluation object comprehensively, respectively from the per share index of target marketing enterprises, profit energy
Power, business growth ability, operation ability, payment of debts and 84 concrete financial index of 6 aspects such as capital structure, cash flow set out, and sentence
Disconnected assessment company generic and the degree of association with relevant industries company thereof.But, undeniable, this 84 financial index institutes
The assessment data not only characteristic dimension constituting is high, redundancy is strong, and often there is also synteny between index and index.Think
Want directly by understanding the relevance of assessment company and industry associated companies thereof to the analysis of these financial index also not
Easy thing.Therefore, the present embodiment proposes a kind of Feature Selection Model based on principal component analysis (PCA), enters assessment data
Row Feature Compression and dimension reduction.PCA is the method for kind of dimensionality reduction, and multiple variables transformations, by means of an orthogonal transformation, are by it
Minority price generalized variable (that is: principal component), wherein each principal component is the linear combination of original variable, between each principal component
Orthogonal, thus these principal components can reflect most information of variable, and information contained non-overlapping copies.By using
Pca model, can obtain assessment data more accurately, and these data had both represented original data characteristics, the data volume that reduces again,
Improve model calculation usefulness, the complexity decreasing system.
For example, descriptive study object (assessment company and the related public affairs of industry thereof can be come with p feature (financial index) herein
Department), use X respectively1,X2,X3,…,XpRepresenting, this p feature constitutes a p dimensional vector X=(X1,X2,X3,…,Xp)T.If this to
The average of amount X is μ, and covariance average is ∑.Assume that X is with the column vector of n sample size variable composition, and μkIts kth of formula
The desired value of individual element, it may be assumed that μk=E (Xk), covariance matrix is defined as:
Linear transformation is carried out to X, it is considered to the linear combination of original variable:
Principal component is uncorrelated d volume linear combination Z1, Z2..., Zp, and Z1It is X1, X2..., XpLinear combination in side
Difference the maximum, Z2It is and Z1Variance the maximum in incoherent linear combination ..., ZpIt is and Z1, Z2..., Zp-1All incoherent
Variance the maximum in linear combination.
Wherein, the above-mentioned basic step carrying out principal component analysis to 84 Corporate Finance achievement datas is as follows:
The first step: choose n the listed company related to assessment company, obtains p=84 finance of this n company respectively
Index, can be obtained matrix X=(x by the initial data of sample estimatesI, j〕n×p, wherein xI, jRepresent the jth item wealth of i-th listed company
Business achievement data.
Second step: for the difference eliminating between every financial index in dimension and the order of magnitude, finance index is carried out
Standardization, obtains normalized matrix.Standardization formula is as follows: y
Wherein, i=1,2 ..., n;J=1,2 ..., p.
3rd step: set up covariance matrix R according to standardized data matrix, is related between the data after reflection standardizes
The statistical indicator of degree in close relations, is worth bigger, illustrates to be necessary to carry out principal component analysis to data.Wherein, battle array RijFor original
Variable xiWith xjCoefficient correlation.RijFor being symmetrical matrix, only need to calculate triangle element or lower triangle element on it, its meter
Calculating formula is:
4th step: obtain characteristic value, principal component contributor rate and accumulative variance contribution ratio according to covariance matrix R, determines main
Composition number.Choose principal component number according to the principle (that is: contribution rate of accumulative total reaches 80%-90%) choosing principal component number to make
It is analyzed for new characteristic variable.
Below, above-mentioned mixed model is described further: hierarchical clustering model is for for analyzing the relevant enterprise degree of association
Kmeans Clustering Model.In above-described embodiment, use the mixed model being made up of Kmeans clustering algorithm and SVM algorithm, use
In the correlation degree obtaining evaluation object the industry.From the point of view of further, first pass through Kmeans Clustering Model to industry and enterprise number
According to clustering, then by the feature structure of this cluster of SVM model learning, in order to the evaluated company of system automatic decision with
Industry is associated the similarity degree of enterprise.
It should be noted that in above-described embodiment, disaggregated model storehouse is each for provide that it stores for intelligent evaluation system
Class model, disaggregated model storehouse is also configured with model base management system, and model base management system is for carrying out mould to the model of storage
Type extracts, accesses, updates and synthetic operation.Wherein, disaggregated model storehouse is provided with special purpose model, routine for similar references object
Analyze model, built-up pattern and temporary pattern;Disaggregated model storehouse is also configured with model base management system, model base management system
It including construction management module, access management module, operational management module, is additionally provided with model dictionary, internal database and outside
Database, for the correlation model managing, calling, configure in disaggregated model storehouse.
In above-described embodiment, use the mixed model being made up of Kmeans clustering algorithm and SVM algorithm, obtain assessment right
As the correlation degree of the industry, and then classification analysis is done to assessment company, below in conjunction with an example, to its further process
Procedure declaration is as follows:
Here, after by doing PCA analysis to the financial index of sample company, can obtain reflecting the main one-tenth of taking-over market change
Variation per minute, then uses the correlation degree of Kmeans Clustering Model research relevant enterprise of the same trade, and beats for this correlation degree
Upper label;SVMs (SVM) model judges in the industry residing for evaluated company according to these labels, if having its phase
As enterprise can be used as assessment reference.
In dynamic evaluation model system, above-mentioned mixed model determines the decision logic of follow-up intelligent evaluation system,
Thus determined what kind of assessment models of selection and is estimated.From the point of view of further, if the mixed model of Kmeans+SVM is judged
There is stronger relevance between some enterprise of evaluated company and place industry, then intelligent evaluation system will select phase
To evaluated company, valuation is carried out to Valuation Modelling;If but Kmeans+SVM mixed model judges that evaluated company not belongs to
The stronger enterprise of the specific relevance of a certain class, then intelligent evaluation system will select absolute Valuation Modelling to enter evaluated company
Row valuation, concrete processing procedure is as follows:
In Kmeans Clustering Model, it is necessary first to the similarity measurement being defined between example, or equivalently, define distance
Tolerance.In the present embodiment, 2 rank norms Min Shi distance (Minkowksi Distance) of use are i.e.: Euclidean distance is measured
Similarity degree between sample data.It is defined as:
By similarity function, the distance between sample point can be measured out.Secondly, it is assumed that extract the collection of initial data
It is combined into Y=(y1, y2..., yi), wherein, i≤p=84.The target of Kmeans cluster is intended to find a division P={C of Z1,
C2..., Ck, make object function minimum:
Here υkIt is expressed as the C that classifieskMean value.
When using Kmeans Clustering Model to analyze the correlation degree of relevant enterprise of the same trade, its concrete processing procedure is such as
Under:
(1) assume that being analyzed, by PCA, the main variables obtaining is Y=(y1, y2..., yi), random taking-up k=2 from Y
Element, as the center of k bunch, k takes 2 and is because here, it is intended that judges and assesses whether related company also exists the degree of association,
Give label+1 if also existing, if do not exist this tagged-1.
(2) first calculate the mean value of an element in Y, calculate the Europe to central point of each element further according to aforementioned formula respectively
Formula distance, by these distance respectively incorporate into similarity minimum bunch in.
(3) according to cluster result, the center of k bunch is recalculated, computing formula:Wherein, rnrCounting
It is 1 when strong point n is classified into k bunch, be otherwise 0.The mode of cross-iteration optimizing can be used to determine υk、rnr, it may be assumed that first solid
Determine υk, to aforementioned formula derivation, whenWhen obtain extreme value rnr.Fix r againnr, to υkSeek local derviation, whenWhen obtain pole
Value υk, repeatedly just can obtain optimal value.
(4) the 3rd step is repeated, until cluster result no longer changes.
(5) result is exported.
Above-described embodiment is obtained in that the division of a degree of similarity of all associated companies by Kmeans cluster, and
Pass through+1 and-1 to be marked.Wherein, if sample yi∈C1, then ri=+1, otherwise, if yi∈C2, then ri=-1.For sample χ
={ yi, ri}.Wish to find ω and ω0Make:
ri(ωTyi+ω0)≥1
Work as ri{ when+1 ,-1}, above formula can be denoted as ∈
Wherein, ρ is certain specific value, it is desirable to it is the bigger the better.But scaling ω, the number of available solution is permissible
It is unlimited number of.In order to obtain unique solution it is necessary to fix ρ | | ω | |=1, so for maximal margin it is necessary to minimize |
| ω ||, therefore, following optimal separating hyperplane can be drawn:
Lagrange multiplier is used to be converted into unconfinement optimization problem, it may be assumed that
Equation is with regard to ω and ω0Minimize, with regard to αi>=0 maximizes, and saddle point goes out to provide understanding value.
Equation is a convex double optimization problem, therefore can use Karush-Kuhn-Tucker its dual problem of condition solution,
Dual problem is with regard to αiMaximize Lp, it is limited to retrain ω and ω0, value when being 0 to its local derviation respectively:
Substituted into formula, drawn dual problem:
So only need to pay close attention to αiTo LdMaximization, its be limited to constraint:Once solve αi's
Occurrence, it can be seen that although they have N number of, but mostly with αiThe form of=0 disappears, and meets α only on a small quantityi>=0 to
The set of amount, constitutes so-called " supporting vector ", and these are supported that vector establishes a classification of problem, thus establish
Sample and mapping relations of classification results, just can calculate Target Enterprise and which sample of assessment according to this mapping function
Similarity relation is there is between enterprise.
Therefore, being clustered by Kmeans, obtaining a mark of all associated companies, this mark dimension one ground determines
Degree of similarity between sample company, then learns this mark by SVM, allows it set up one between sample and classification
Individual rational mapping, this mapping relations can help to judge, whether specific sample belongs in a stronger group of relevance.This
Sample does and is advantageous in that: when when carrying out value of stock right assessment to company, can first pass through SVM algorithm and determine whether and mesh
The similar company of mark Comments on Companies'Financial Condition.If there is similar company, relative Valuation Modelling will be used for the value of stock right
Assessment, if there is not similar company, then absolute Valuation Modelling just can be used.
As the optional embodiment of one, in above-described embodiment, evaluation object is the value of stock right of marketing enterprises, assessment
The identity information of object is CompanyName or stock code.Above-mentioned intelligent evaluation system uses a kind of condition based on binary tree to sentence
The selection to Valuation Modelling of breaking makes inferences screening, and above-mentioned intelligent evaluation system can first can be according to the computing of disaggregated model bunch
As a result, judge whether evaluated company has and there is higher similarity with associated companies of the same trade, and in Valuation Modelling storehouse
In choose optimal models equity assessment carried out to company to be assessed.
As the optional embodiment of one, in above-described embodiment, relative valuation mode include p/e ratio (P/E) model,
HSBC (P/B) model and city's pin rate (P/S) model;Intelligent evaluation system is according to designing based on knowwhy and expertise
Assessment parameter, chooses corresponding optimum evaluation model and carries out share price value assessment to the evaluation object belonging to gregarious sample.
As the optional embodiment of one, in above-described embodiment, absolute valuation mode includes that free cash flow discounts (NPV)
Model, option on physicals (ROpt.) Valuation Modelling;Intelligent evaluation system is according to the assessment designing based on knowwhy and expertise
Parameter, chooses corresponding optimum evaluation model and carries out share price value assessment to the evaluation object belonging to outliers.
As the optional embodiment of one, in above-described embodiment, intelligent evaluation system arrange screening conditions for based on
The reasoning from logic principle that knowwhy and expertise design, is packaged in intelligent evaluation system with code form, and presses suitable
Sequence encodes, and can be normalized, systematized call.
The various embodiments described above are by introducing the technology such as information search, data mining, expertise reasoning, for listed company
The value of stock right, establishes a set of dynamic evaluation model system based on IDSS and dynamic assessment method.Use mathematical logic and
Algorithm is deduced, and builds the Evaluation Platform of a circulation, in conjunction with the relative valuation with features such as simplification, facilitation and ageings
Method with there is the absolute technique of estimation of feature such as objectivity, accuracy and completeness, formed complete set Intelligent Dynamic evaluation system.
So both evade subjective error produced by traditional Manpower Estimate, reduce again the one-tenth that evaluation process is shone to greatest extent
Manpower is freed from heavy evaluation work by this loss.
It should be noted that for aforesaid embodiment of the method, in order to be briefly described, therefore it is all expressed as a series of
Combination of actions, but those skilled in the art should know, and the present invention is not limited by described sequence of movement, because depending on
According to the present invention, some step can use other orders or carry out simultaneously.Secondly, those skilled in the art also should know,
Embodiment described in this description belongs to preferred embodiment, and involved action is not necessarily essential to the invention.
System embodiment
For realizing said method, the present invention proposes a kind of dynamic evaluation model system, with reference to the group of this system shown in Fig. 2
Becoming block diagram, this dynamic evaluation model system includes consisting of:
1) man-machine interactive system, based on single interface encapsulation, has visualization appraisal framework, be provided with human-computer interaction interface and
Result output display interface;
2) dynamic data base system, based on teledata search pattern, has open dynamic online database, is used for root
According to the identity information of evaluation object, set screening conditions, searching route and search keyword search and obtain with regard to assessment
The html source code at the target data place of object, and Data Analysis and reconstruct are carried out to html source code;
3) disaggregated model storehouse system, for the target data obtaining according to dynamic data base system, extraction can reflect to be commented
Estimate the structural characterization data of object, structural characterization data is identified and classification analysis computing;
4) intelligent evaluation system, for according to the operation result of disaggregated model storehouse system, it is judged that evaluation object be belonging to from
Group's sample still falls within gregarious sample, and chooses corresponding optimum evaluation model and be estimated evaluation object;Work as evaluation object
When belonging to gregarious sample, for according to the assessment parameter of classification analysis result and setting, call in the system of disaggregated model storehouse with
The assessment models that evaluation object matches, uses relative valuation mode, is estimated evaluation object, and exports assessment result;
And, when evaluation object belongs to outliers, for the assessment parameter according to classification analysis result and setting, use and definitely estimate
Value mode, is estimated to evaluation object, and exports assessment result.
In above-described embodiment, man-machine interactive system uses the visualization appraisal framework (Single of single interface encapsulation
Interface Package Visualization Framework, SIPVF), so so that OO company stock
Power value assessment process becomes short and sweet.For example, in traditional stock price value assessment, appraiser's not only GPRS
The related financial savvy of assessment is theoretical with enterprise value, but also will before assessment to the concrete situation of Suo Ping company (such as: company
Industrial nature, the financial index of company, the management state of type of industry, the economic environment etc. in market) carry out Data Collection with
Arrange, and make concrete analysis according to these data;Single interface encapsulation technology that SIPVF uses then is evaded to greatest extent
Such problem.Single interface encapsulation refers to, we by the treatment mechanism of whole dynamic evaluation (include: dynamic data base system,
Disaggregated model storehouse system, foundation and the utilization of intelligent evaluation system) Hide All in internal system, only reserved one external defeated
Inbound port.Appraiser only need to be to the title of system input evaluation object (listed company) or stock code, dynamic evaluation model
System just can complete, from assessment Data Collection, to arrive data classification analysis, then choose to finance model, and comment to the value of stock right
Estimate, until the overall process of assessment result output.Therefore, use this OO single interface encapsulation technology, system can be assessed
Convenient to operate, also corresponding evaluation object can be made reasonably estimating even if helping those not know about the personnel of equity Assessment theory
Value.
Additionally, the SIPVF of dynamic evaluation model system uses Visualization Framework technology, help to use dynamic data base system
Appraiser can clearly, be visually known in dynamic evaluation model system, the assessment result of the value of stock right is by which
The impact of a little material elementses and influence degree have much, thus aid decision making personnel are in follow-up Analysis on Enterprise Strategy, do
Go out corresponding decision-making to judge.In the present embodiment, Visualization Framework be use affine projection (Affine Transformation),
The modes such as dimensionality reduction (Re-dimensional) analyze the global information of multidimensional data, and this makes the decision-maker can be more directly perceived
Carry out data analysis easily.
As the optional embodiment of one, in above-described embodiment, dynamic data base system includes: open long-range
Line database, local data base and data base management system;Dynamic data base system uses open crawler algorithm as building storehouse
Engine.
Wherein, dynamic data base system uses a kind of online data warehouse based on open crawler algorithm.This number
It is according to the mentality of designing in warehouse: the present embodiment uses a kind of open remote online data warehouse, is commenting as value
When estimating, different from traditional local data warehouse, it is not required to manually go to collect the data required for assessment, but use one
Pre-designed crawler algorithm can at the open remote server port of the one (or several) specified (such as: large-scale synthesis
The API port of class financial web site) search corresponding assessment data, then the destructuring that will be crawled on networking by regular expression
Data Analysis is structural data, finally by locally downloading for the structural data an obtaining volatile data base, as stock
The data of power valuation are supported.
As the optional embodiment of one, in above-described embodiment, disaggregated model storehouse system include disaggregated model bunch and
Model base management system, disaggregated model bunch provides its each class model storing for intelligent evaluation system, and model base management system is used
In to storage model carry out model extraction, accesss, renewal and synthetic operation.Disaggregated model storehouse system is used for calling and pre-builds
Disaggregated model bunch in Data Analysis Model, the structural characterization data of evaluation object is extracted;And, it is used for using pre-
The mixed model first set up carries out classification analysis to evaluation object.
From the point of view of further, in above-described embodiment, model involved in disaggregated model bunch is broadly divided into two big classes:
1st, Data Analysis Model
Data Analysis Model includes for carrying out Feature Compression and the principal component analysis of dimension reduction to evaluation object
(PCA) model;Mixed model includes: hierarchical clustering model and SVMs (SVM) model, hierarchical clustering model is used for identifying
Evaluation object and the correlation degree of similar references object the hierarchical clustering model stamping correlation tag for this correlation degree;
SVMs (SVM) model can same as assessment reference for judging whether evaluation object can match according to correlation tag
Class references object, if can match can be as the similar references object of assessment reference for evaluation object, then evaluation object belongs to conjunction
Group's sample, otherwise evaluation object belongs to outliers.
From the point of view of further, pca model is used for being estimated the feature extraction of data, analyzes various financial index to company's valency
The impact of value, and how compressed by index and form new index, these new indexs can reflect the overall wealth of company
Business situation can reduce again the complexity of analysis.Hierarchical clustering model and SVMs (SVM) model, be used for looking forward to assessment target
Industry is classified, and will do a category division to assessment target enterprise and relevant enterprise thereof, and this category division is conducive to help
Dynamic evaluation system judges it is to use relative technique of estimation or absolute technique of estimation during equity valuation.
2nd, it is used for value of stock right assessment models
Assessment models includes: p/e ratio (P/E) model in the first model library, HSBC (P/B) model, city pin rate (P/
S) model;And, the free cash flow in the second model library is discounted (NPV) model and option on physicals (ROpt.);Wherein, described
The assessment parameter that in first model library and described second model library, the Call Condition of each model is arranged by described intelligent evaluation system
Determine.Wherein:
(1) with p/e ratio (P/E) model based on relative technique of estimation theory, HSBC (P/B) model and city's pin rate (P/S)
Model, the target enterprise shipping as belonging to gregarious sample is carried out equity valuation by them;
(2) with the option on physicals Valuation Modelling based on absolute technique of estimation theory, it will be used for the mark belonging to outliers
Enterprise carry out the assessment of share price value.
It is emphasized that in above-described embodiment, intelligent evaluation system is used for calling, manages described dynamic data base system
And disaggregated model storehouse system, and to carry out information mutual with described man-machine interactive system.For example, intelligent evaluation system is useful for
Value of stock right assessment models such as p/e ratio (P/E) model, HSBC (P/B) model, city's pin rate (P/S) model, and net cash
Stream is discounted (NPV) model and option on physicals (ROpt.), arranges the Call Condition of each model, and Call Condition can be by arranging
Assessment parameter determines.
Functionally dividing, above-mentioned intelligent evaluation management system mays include: body field module, for providing to related reason
The shared ideas model of opinion knowledge and Formal Specification explanation;Reasoning from logic module: be used for providing correlation theory to patrol with corresponding
The matching way of the relation of collecting;Task calculating module: for carrying out according to the Valuation Modelling determining calculating and result of calculation being exported.
From the point of view of further, in the module of body field, build field set, including knowwhy territory and relation domain.Wherein, originally
Body is the shared ideas model to relevant knowledge and the explanation of clear and definite Formal Specification, it provides in intelligent evaluation management system
Basic terms (knowwhy atom) and relation, and utilize these basic terms and relation constitution theory knowledge rule for extent and
Complex definitions.
From the point of view of further, in reasoning from logic module, reasoning from logic can provide mating of a kind of knowwhy and logical relation
Mode, by by the knowwhy symbolism in knowwhy territory, encoding body, makes knowwhy and corresponding logic
Relation is mated, so that it is determined that the logical relation of all knowwhies in overall domain body.
From the point of view of further, in task calculating module, determine the target output of task, calculated by reasoning from logic relation and be
System output such that it is able to provide the final output of intelligent evaluation system.
As the optional embodiment of one, in above-described embodiment, intelligent evaluation system arrange screening conditions for based on
The reasoning from logic principle that knowwhy and expertise design, is packaged in intelligent evaluation system with code form, and presses suitable
Sequence encodes, and can be normalized, systematized call.
It is pointed out that the dynamic evaluation model system in above-described embodiment is corresponding with above-mentioned dynamic assessment method,
Concrete implementation process can refer to preceding method embodiment.Owing to any of the above-described kind of dynamic assessment method has above-mentioned technology effect
Really, therefore, this dynamic evaluation model system also should possess corresponding technique effect, its specific implementation process and above-described embodiment class
Seemingly, hereby do not repeat.
Obviously, those skilled in the art should be understood that in the dynamic evaluation model system of the above-mentioned embodiment of the present invention
Each module or dynamic assessment method in each step can realize by general computing device, they can concentrate on single
Computing device on, or be distributed on the network that multiple computing device is formed, alternatively, they can by computing device
The program code performing realizes, it is thus possible to be stored in being performed by computing device in storage device, or by it
Be fabricated to each integrated circuit modules respectively, or the multiple module in them or step are fabricated to single integrated circuit mould
Block realizes.So, the present invention is not restricted to the combination of any specific hardware and software.Described storage device is non-volatile depositing
Reservoir, such as: ROM/RAM, flash memory, magnetic disc, CD etc..
The foregoing is only embodiments of the invention, not in order to limit the present invention, all spirit in the present invention and
Within principle, any modification, equivalent substitution and improvement etc. made, should be included within the scope of the present invention.
Claims (10)
1. a dynamic assessment method, it is characterised in that include:
Pre-build the dynamic evaluation model system based on intelligence decision support system system (IDSS), described dynamic evaluation model system bag
Include: based on the open dynamic data base system of teledata search pattern, disaggregated model storehouse system, intelligent evaluation system, with
And the man-machine interactive system of the visualization appraisal framework based on single interface encapsulation;
Described intelligent evaluation system obtains the identity information of evaluation object by man-machine interactive system;
Calling described dynamic data base system, described dynamic data base system is according to the identity information of described evaluation object, described
Screening conditions, searching route and search keyword search that intelligent evaluation system is arranged simultaneously obtain mesh with regard to described evaluation object
The html source code at mark data place, and Data Analysis and reconstruct are carried out to described html source code;
Call described disaggregated model storehouse system, the target that described disaggregated model storehouse system obtains according to described dynamic data base system
Data, extract the structural characterization data that can reflect described evaluation object, are identified described structural characterization data and classify
Analytic operation;
Described intelligent evaluation system is according to the operation result of described disaggregated model storehouse system, it is judged that described evaluation object be belonging to from
Group's sample still falls within gregarious sample, and chooses corresponding optimum evaluation model and be estimated described evaluation object;If it is described
Evaluation object belongs to gregarious sample, then the assessment parameter according to described classification analysis result and setting, in described disaggregated model storehouse
System is called the assessment models matching with described evaluation object, uses relative valuation mode, described evaluation object is carried out
Assessment, and export assessment result;If described evaluation object belongs to outliers, then according to described classification analysis result and setting
Assessment parameter, uses absolute valuation mode, is estimated described evaluation object, and exports assessment result.
2. dynamic assessment method according to claim 1, it is characterised in that described dynamic data base system is according to institute's commentary
Estimate identity information, the screening conditions of described intelligent evaluation system setting, searching route and the search keyword search of object and obtain
Take the html source code at the target data place with regard to described evaluation object, and Data Analysis and reconstruct are carried out to described html source code
Including:
Select evaluation object according to the identity information of described evaluation object, obtain corresponding screening conditions, the search pre-setting
Path and search keyword;
Described dynamic data base system uses the remote online Database pattern based on open crawler algorithm, when institute's commentary
Estimate object selected after, assessed according to evaluation object relevant information and default URL site search by open crawler algorithm
Desired data, accesses one or more remote server port in the way of keyword response, collects and download corresponding target
The html source code at data place;Wherein, described remote server port is stored in described dynamic data base system, passes through URL
The mode called accesses target pages;
Resolve by way of regular expression character string tagsort, download described html source code to home server, thus
The dynamic interim remote online database of form one;Wherein, the keyword phase of described regular expression and described target data
Coupling.
3. dynamic assessment method according to claim 1 and 2, it is characterised in that described disaggregated model storehouse system is according to institute
State dynamic data base system, extract the structural characterization data that can reflect described evaluation object, described structural characterization data is entered
Row identifies and classification analysis includes:
The structural characterization data to described evaluation object for the Data Analysis Model pre-building is used to extract;Wherein, described
Data Analysis Model includes for principal component analysis (PCA) mould carrying out Feature Compression and dimension reduction to described evaluation object
Type;
The mixed model pre-building is used to carry out classification analysis to described evaluation object;Wherein, described mixed model includes: layer
Secondary Clustering Model and SVMs (SVM) model, described hierarchical clustering model is used for identifying described evaluation object and similar ginseng
Examine the correlation degree of object the hierarchical clustering model stamping correlation tag for this correlation degree;SVMs (SVM) mould
Type can be as the similar reference pair of assessment reference for judging whether described evaluation object can match according to described correlation tag
As if can match can be as the similar references object of assessment reference for described evaluation object, then described evaluation object belongs to conjunction
Group's sample, otherwise described evaluation object belongs to outliers.
4. dynamic assessment method according to claim 3, it is characterised in that:
Described disaggregated model storehouse is for providing its each class model storing for described intelligent evaluation system, and described disaggregated model storehouse is also
Being configured with model base management system, described model base management system for carrying out model extraction, access, updating to the model of storage
And synthetic operation;
Wherein, described disaggregated model storehouse for similar references object be provided with special purpose model, routine analysis model, built-up pattern with
And temporary pattern;Described disaggregated model storehouse is also configured with model base management system, and described model base management system includes construction pipe
Reason module, access management module, operational management module, be additionally provided with model dictionary, internal database and external data base, be used for
The correlation model managing, calling, configure in described disaggregated model storehouse.
5. the dynamic assessment method according to any one of Claims 1-4, it is characterised in that described evaluation object is target
The value of stock right of enterprise, the identity information of described evaluation object is CompanyName or stock code;
Described relative valuation mode includes p/e ratio (P/E) model, HSBC (P/B) model and city's pin rate (P/S) model;Described
Intelligent evaluation system, according to the assessment parameter designing based on knowwhy and expertise, chooses corresponding optimum evaluation model pair
The evaluation object belonging to gregarious sample carries out share price value assessment;
Described absolute valuation mode includes that free cash flow is discounted (NPV) model, option on physicals (ROpt.) Valuation Modelling;Described intelligence
System of assessing, according to the assessment parameter designing based on knowwhy and expertise, chooses corresponding optimum evaluation model to genus
Carry out share price value assessment in the evaluation object of outliers;
Wherein, described hierarchical clustering model is for for analyzing the Kmeans Clustering Model of the relevant enterprise degree of association.
6. the dynamic assessment method according to any one of Claims 1-4, it is characterised in that described intelligent evaluation system sets
The screening conditions put are the reasoning from logic principle designing based on knowwhy and expertise, are packaged in institute with code form
State intelligent evaluation system, and encode in order, can be normalized, systematized call.
7. a dynamic evaluation model system, it is characterised in that include:
Man-machine interactive system, based on single interface encapsulation, has visualization appraisal framework, is provided with human-computer interaction interface and result is defeated
Go out display interface;
Dynamic data base system, based on teledata search pattern, has open dynamic online database, for according to described
The identity information of evaluation object, set screening conditions, searching route and search keyword search simultaneously obtain with regard to institute's commentary
Estimate the html source code at the target data place of object, and Data Analysis and reconstruct are carried out to described html source code;
Disaggregated model storehouse system, for the target data obtaining according to described dynamic data base system, extraction can reflect described
The structural characterization data of evaluation object, is identified to described structural characterization data and classification analysis computing;
Intelligent evaluation system, for the operation result according to described disaggregated model storehouse system, it is judged that described evaluation object is belonging to
Outliers still falls within gregarious sample, and chooses corresponding optimum evaluation model and be estimated described evaluation object;Work as institute
State evaluation object when belonging to gregarious sample, for the assessment parameter according to described classification analysis result and setting, in described classification
Model-base management system is called the assessment models matching with described evaluation object, uses relative valuation mode, right to described assessment
As being estimated, and export assessment result;And, when described evaluation object belongs to outliers, for according to described classification
The assessment parameter of analysis result and setting, uses absolute valuation mode, is estimated described evaluation object, and exports assessment knot
Really.
8. dynamic evaluation model system according to claim 7, it is characterised in that:
Described dynamic data base system includes: open remote online database, local data base and data base management system;
Described dynamic data base system uses open crawler algorithm as building storehouse engine;And/or,
Described disaggregated model storehouse system includes disaggregated model bunch and model base management system, and described disaggregated model bunch is described intelligence
System of assessing provides its each class model storing, and described model base management system carries for carrying out model to the model of storage
Take, access, update and synthetic operation;Described disaggregated model storehouse system is for using the Data Analysis Model that pre-builds to described
The structural characterization data of evaluation object carries out extracting and for using the mixed model pre-building to enter described evaluation object
Row classification analysis;
Wherein, described disaggregated model bunch includes Data Analysis Model, described Data Analysis Model include for described assessment right
Principal component analysis (PCA) model carrying out Feature Compression and dimension reduction of elephant;Described mixed model includes: hierarchical clustering model
With SVMs (SVM) model, described hierarchical clustering model is for identifying the pass of described evaluation object and similar references object
Connection degree the hierarchical clustering model stamping correlation tag for this correlation degree;SVMs (SVM) model is used for basis
Described correlation tag judge described evaluation object whether can match can as the similar references object of assessment reference, if described
Evaluation object can match can be as the similar references object of assessment reference, then described evaluation object belongs to gregarious sample, otherwise
Described evaluation object belongs to outliers;
Described disaggregated model bunch also includes value of stock right assessment models, and described value of stock right assessment models includes: the first model library
In p/e ratio (P/E) model, HSBC (P/B) model, city's pin rate (P/S) model;And, the net cash in the second model library
Stream is discounted (NPV) model and option on physicals (ROpt.);Wherein, each mould in described first model library and described second model library
The assessment parameter that the Call Condition of type is arranged by described intelligent evaluation system determines.
9. the dynamic evaluation model system according to claim 7 or 8, it is characterised in that described intelligent evaluation system is used for
Call, manage described dynamic data base system and disaggregated model storehouse system, and to carry out information mutual with described man-machine interactive system;
Described intelligent evaluation system includes: body field module, for providing the shared ideas model to correlation theory and form
Change specification explanation;Reasoning from logic module: for providing the matching way of correlation theory and corresponding logical relation;Task computation mould
Block: for carrying out according to the Valuation Modelling determining calculating and result of calculation being exported.
10. dynamic evaluation model system according to claim 9, it is characterised in that:
The screening conditions that described intelligent evaluation system is arranged are the reasoning from logic designing based on knowwhy and expertise
Principle, is packaged in described intelligent evaluation system with code form, and encodes in order, can be normalized, systematized tune
With.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610416226.9A CN106067094A (en) | 2016-06-14 | 2016-06-14 | A kind of dynamic assessment method and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610416226.9A CN106067094A (en) | 2016-06-14 | 2016-06-14 | A kind of dynamic assessment method and system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106067094A true CN106067094A (en) | 2016-11-02 |
Family
ID=57420333
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610416226.9A Pending CN106067094A (en) | 2016-06-14 | 2016-06-14 | A kind of dynamic assessment method and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106067094A (en) |
Cited By (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107145435A (en) * | 2017-05-27 | 2017-09-08 | 北京仿真中心 | A kind of assessment of performance system and method based on B/S frameworks |
CN107248024A (en) * | 2017-05-19 | 2017-10-13 | 武汉理工大学 | The method of assessment submarine student's simulated training result based on SVM algorithm |
CN107644290A (en) * | 2017-09-05 | 2018-01-30 | 河北工程大学 | A kind of patented technology Life Cycle Analysis |
CN108121750A (en) * | 2016-11-30 | 2018-06-05 | 西门子公司 | A kind of model treatment method, apparatus and machine readable media |
CN108573451A (en) * | 2017-03-09 | 2018-09-25 | 派衍信息科技(苏州)有限公司 | A kind of security type diversification valuation processing system |
CN108960772A (en) * | 2018-06-27 | 2018-12-07 | 北京窝头网络科技有限公司 | Enterprise's evaluation householder method and system based on deep learning |
CN109118085A (en) * | 2018-08-14 | 2019-01-01 | 石榴籽科技有限公司 | A kind of building trade risk management and control system and method Internet-based |
CN109544337A (en) * | 2018-11-15 | 2019-03-29 | 北京心流慧估科技有限公司 | A kind of equity estimation method |
CN109635833A (en) * | 2018-10-30 | 2019-04-16 | 银河水滴科技(北京)有限公司 | A kind of image-recognizing method and system based on cloud platform and model intelligent recommendation |
TWI677843B (en) * | 2017-09-15 | 2019-11-21 | 群益金鼎證券股份有限公司 | Intelligent cluster suggestion system and method |
CN110928861A (en) * | 2018-09-18 | 2020-03-27 | 上汽通用汽车有限公司 | Auxiliary analysis and evaluation method and system for vehicle road noise |
CN110941246A (en) * | 2019-10-22 | 2020-03-31 | 杭州电子科技大学 | HMI message shunting scheduling method, storage medium and device |
CN111144677A (en) * | 2018-11-06 | 2020-05-12 | 北京京东振世信息技术有限公司 | Efficiency evaluation method and efficiency evaluation system |
CN111241483A (en) * | 2019-01-11 | 2020-06-05 | 深圳联合产权交易所股份有限公司 | Resource value evaluation processing method based on cloud platform and related products |
CN112085331A (en) * | 2020-08-04 | 2020-12-15 | 广东省科学技术情报研究所 | Research and development mechanism dynamic monitoring method and system based on big data |
CN112508440A (en) * | 2020-12-18 | 2021-03-16 | 深圳市赛为智能股份有限公司 | Data quality evaluation method and device, computer equipment and storage medium |
CN112995201A (en) * | 2019-01-11 | 2021-06-18 | 深圳联合产权交易所股份有限公司 | Resource value evaluation processing method based on cloud platform and related device |
CN113112380A (en) * | 2021-05-12 | 2021-07-13 | 北京大学 | Intellectual property service value evaluation system |
CN113159844A (en) * | 2021-04-20 | 2021-07-23 | 上海外国语大学 | Intelligent advertisement evaluation method and system based on eyeball trajectory tracking |
-
2016
- 2016-06-14 CN CN201610416226.9A patent/CN106067094A/en active Pending
Cited By (26)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108121750A (en) * | 2016-11-30 | 2018-06-05 | 西门子公司 | A kind of model treatment method, apparatus and machine readable media |
CN108121750B (en) * | 2016-11-30 | 2022-07-08 | 西门子公司 | Model processing method and device and machine readable medium |
CN108573451A (en) * | 2017-03-09 | 2018-09-25 | 派衍信息科技(苏州)有限公司 | A kind of security type diversification valuation processing system |
CN107248024A (en) * | 2017-05-19 | 2017-10-13 | 武汉理工大学 | The method of assessment submarine student's simulated training result based on SVM algorithm |
CN107145435A (en) * | 2017-05-27 | 2017-09-08 | 北京仿真中心 | A kind of assessment of performance system and method based on B/S frameworks |
CN107644290A (en) * | 2017-09-05 | 2018-01-30 | 河北工程大学 | A kind of patented technology Life Cycle Analysis |
TWI677843B (en) * | 2017-09-15 | 2019-11-21 | 群益金鼎證券股份有限公司 | Intelligent cluster suggestion system and method |
CN108960772A (en) * | 2018-06-27 | 2018-12-07 | 北京窝头网络科技有限公司 | Enterprise's evaluation householder method and system based on deep learning |
CN109118085A (en) * | 2018-08-14 | 2019-01-01 | 石榴籽科技有限公司 | A kind of building trade risk management and control system and method Internet-based |
CN110928861B (en) * | 2018-09-18 | 2023-11-17 | 上汽通用汽车有限公司 | Auxiliary analysis and evaluation method and system for vehicle road noise |
CN110928861A (en) * | 2018-09-18 | 2020-03-27 | 上汽通用汽车有限公司 | Auxiliary analysis and evaluation method and system for vehicle road noise |
CN109635833A (en) * | 2018-10-30 | 2019-04-16 | 银河水滴科技(北京)有限公司 | A kind of image-recognizing method and system based on cloud platform and model intelligent recommendation |
CN111144677A (en) * | 2018-11-06 | 2020-05-12 | 北京京东振世信息技术有限公司 | Efficiency evaluation method and efficiency evaluation system |
CN111144677B (en) * | 2018-11-06 | 2023-11-07 | 北京京东振世信息技术有限公司 | Efficiency evaluation method and efficiency evaluation system |
CN109544337A (en) * | 2018-11-15 | 2019-03-29 | 北京心流慧估科技有限公司 | A kind of equity estimation method |
CN112995201B (en) * | 2019-01-11 | 2022-07-12 | 深圳联合产权交易所股份有限公司 | Resource value evaluation processing method based on cloud platform and related device |
CN111241483B (en) * | 2019-01-11 | 2020-10-30 | 深圳联合产权交易所股份有限公司 | Resource value evaluation processing method based on cloud platform and related products |
CN111241483A (en) * | 2019-01-11 | 2020-06-05 | 深圳联合产权交易所股份有限公司 | Resource value evaluation processing method based on cloud platform and related products |
CN112995201A (en) * | 2019-01-11 | 2021-06-18 | 深圳联合产权交易所股份有限公司 | Resource value evaluation processing method based on cloud platform and related device |
CN110941246B (en) * | 2019-10-22 | 2021-03-16 | 杭州电子科技大学 | HMI message shunting scheduling method, storage medium and device |
CN110941246A (en) * | 2019-10-22 | 2020-03-31 | 杭州电子科技大学 | HMI message shunting scheduling method, storage medium and device |
CN112085331A (en) * | 2020-08-04 | 2020-12-15 | 广东省科学技术情报研究所 | Research and development mechanism dynamic monitoring method and system based on big data |
CN112508440A (en) * | 2020-12-18 | 2021-03-16 | 深圳市赛为智能股份有限公司 | Data quality evaluation method and device, computer equipment and storage medium |
CN112508440B (en) * | 2020-12-18 | 2024-06-07 | 深圳市赛为智能股份有限公司 | Data quality evaluation method, device, computer equipment and storage medium |
CN113159844A (en) * | 2021-04-20 | 2021-07-23 | 上海外国语大学 | Intelligent advertisement evaluation method and system based on eyeball trajectory tracking |
CN113112380A (en) * | 2021-05-12 | 2021-07-13 | 北京大学 | Intellectual property service value evaluation system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106067094A (en) | A kind of dynamic assessment method and system | |
US10977293B2 (en) | Technology incident management platform | |
CN102508860A (en) | Data mining method based on XBRL (extensible business reporting language) embodiment document | |
CN109543925A (en) | Risk Forecast Method, device, computer equipment and storage medium based on machine learning | |
Pan et al. | Clustering of designers based on building information modeling event logs | |
CN112417176B (en) | Method, equipment and medium for mining implicit association relation between enterprises based on graph characteristics | |
CN107292744A (en) | Investment Trend analysis method and its system based on machine learning | |
Xu et al. | Novel key indicators selection method of financial fraud prediction model based on machine learning hybrid mode | |
CN112819308A (en) | Head enterprise identification method based on bipartite convolutional neural network | |
CN115794803B (en) | Engineering audit problem monitoring method and system based on big data AI technology | |
CN101546290A (en) | Method for improving accuracy of quality forecast of class hierarchy in object-oriented software | |
CN104933621A (en) | Big data analysis system and method for guarantee ring | |
CN115358481A (en) | Early warning and identification method, system and device for enterprise ex-situ migration | |
Si | Construction and application of enterprise internal audit data analysis model based on decision tree algorithm | |
CN117764724A (en) | Intelligent credit rating report construction method and system | |
Yu et al. | Computational intelligent data analysis for sustainable development | |
Jeyaraman et al. | Practical Machine Learning with R: Define, build, and evaluate machine learning models for real-world applications | |
Lin et al. | Applying the random forest model to forecast the market reaction of start-up firms: case study of GISA equity crowdfunding platform in Taiwan | |
Chang et al. | Automated feature engineering for fraud prediction in online credit loan services | |
Hou | Financial Abnormal Data Detection System Based on Reinforcement Learning | |
Sridhar et al. | Effect of rate of change of stock prices with news sentiment analysis | |
CN113379211A (en) | Block chain-based logistics information platform default risk management and control system and method | |
CN113379212A (en) | Block chain-based logistics information platform default risk assessment method, device, equipment and medium | |
Zhou et al. | [Retracted] An Improved Data Mining Model for Predicting the Impact of Economic Fluctuations | |
CN112699245A (en) | Construction method and device and application method and device of budget management knowledge graph |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20161102 |