CN109583940A - Cell source of houses value parameter estimation method and device - Google Patents
Cell source of houses value parameter estimation method and device Download PDFInfo
- Publication number
- CN109583940A CN109583940A CN201811289888.XA CN201811289888A CN109583940A CN 109583940 A CN109583940 A CN 109583940A CN 201811289888 A CN201811289888 A CN 201811289888A CN 109583940 A CN109583940 A CN 109583940A
- Authority
- CN
- China
- Prior art keywords
- price
- houses
- average price
- source
- data
- 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 46
- 238000013528 artificial neural network Methods 0.000 claims abstract description 110
- 238000012549 training Methods 0.000 claims description 63
- 238000012360 testing method Methods 0.000 claims description 49
- 238000004590 computer program Methods 0.000 claims description 27
- 238000012797 qualification Methods 0.000 claims description 13
- 238000012545 processing Methods 0.000 claims description 7
- 230000003993 interaction Effects 0.000 claims description 3
- 241000208340 Araliaceae Species 0.000 claims description 2
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 claims description 2
- 235000003140 Panax quinquefolius Nutrition 0.000 claims description 2
- 235000008434 ginseng Nutrition 0.000 claims description 2
- 238000010801 machine learning Methods 0.000 abstract 1
- 230000006870 function Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 230000002159 abnormal effect Effects 0.000 description 4
- 238000003197 gene knockdown Methods 0.000 description 4
- 238000013461 design Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 230000004913 activation Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 201000004569 Blindness Diseases 0.000 description 1
- 241001269238 Data Species 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000005036 nerve Anatomy 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0278—Product appraisal
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/16—Real estate
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Strategic Management (AREA)
- General Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Tourism & Hospitality (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Economics (AREA)
- Finance (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Primary Health Care (AREA)
- Entrepreneurship & Innovation (AREA)
- Human Resources & Organizations (AREA)
- Game Theory and Decision Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
This application involves machine learning and field of neural networks, for handling cell source of houses data, and in particular to a kind of cell source of houses value parameter estimation method, device, computer equipment and storage medium.Method includes: to obtain listed average price, conclusion of the business average price and the initial market average price of each preset time period of the Target cell source of houses；The prediction of the grey multivariable degree of association is carried out to listed average price, conclusion of the business average price and initial market average price, obtains price associated data；Listed average price, conclusion of the business average price and price associated data are inputted into preset wavelet neural network, obtain the prediction average price of the Target cell source of houses.The application can carry out accurate valuation by average value parameter of the preset wavelet neural network to the Target cell source of houses according to the cell degree of association between the price data and price data of the conclusion of the business source of houses, more objective and accurate compared to the subjectivity appraisal of appraiser.
Description
Technical field
This application involves field of information processing, more particularly to a kind of cell value parameter estimation method and device.
Background technique
With the development of economy with the propulsion of urbanization, people's lives level also constantly promoted.In city, cell
Refer to based on residence building and is formed equipped with commercial network, culture and education, amusement, greening, public and communal facility etc.
Resident living area of certain scale.The main body of cell is resident's building, and a cell generally comprises several residents and lives
Room, community resident house are the focuses of present investment in property.
However traditional house property estimation method need to consider with the knock-down price of house type identical in cell the valuation in single set house
Lattice can not carry out accurate valuation to house to be assessed under conditions of no conclusion of the business or listed data can be for reference.
Summary of the invention
Based on this, it is necessary in view of the above technical problems, provide a kind of cell source of houses value parameter estimation method, device,
Computer equipment and storage medium.
A kind of cell source of houses value parameter estimation method, which comprises
Obtain listed average price, conclusion of the business average price and the initial market average price of each preset time period of the Target cell source of houses；
Grey multivariable is carried out to the listed average price, the conclusion of the business average price and the initial market average price of the cell source of houses
Degree of association prediction, obtains price associated data；
The listed average price, the conclusion of the business average price and the price associated data are inputted into preset Wavelet Neural Network
Network obtains the prediction average value parameter of the Target cell source of houses, and the preset wavelet neural network is with Target cell room
The historical price data of each historical time section in source are obtained as training data；
Obtain the target source of houses each appraisal influence factor, according to the prediction average value parameter of the Target cell source of houses with
And each appraisal influence factor of the target source of houses estimates that the value of the target source of houses is joined by presetting half parameter quantile regression
Number, the default half parameter quantile regression are based on each appraisal influence factor and the prediction average value parameter to institute
The influence for stating the value parameter of the target source of houses is established.
The listed average price for obtaining each preset time period of the Target cell source of houses, conclusion of the business in one of the embodiments,
Before average price and initial market average price, further includes:
Obtain the listed price data and concluded price data of each preset time period of the Target cell source of houses；According to described
Listed price data, the listed average price for obtaining each preset time period of the Target cell source of houses are obtained according to the concluded price data
Obtain the conclusion of the business average price of each preset time period of the Target cell source of houses.
It is described according to the listed price data in one of the embodiments, it is each default to obtain the Target cell source of houses
The listed average price of period, according to the concluded price data, the conclusion of the business for obtaining each preset time period of the Target cell source of houses is equal
Before valence, further includes:
Processing is filtered to the listed price data and the concluded price data based on mean filter principle.
It is described by the listed average price, the conclusion of the business average price and the price incidence number in one of the embodiments,
According to the preset wavelet neural network of input, before the prediction average value parameter for obtaining the Target cell source of houses, further includes:
Obtain the historical price data of each historical time section of the Target cell source of houses, the historical price data packet
Include listed average price, conclusion of the business average price, price associated data and initial market average price, the listed average price, conclusion of the business average price and valence
Lattice associated data is input data, and the initial market average price is prediction result；
The historical price data are divided into training set and test set at random；
The training set is inputted into initial wavelet neural network, initial wavelet neural network is carried out by gradient coaching method
Training；
The initial wavelet neural network after the training is assessed by the test set, obtains assessment result；
When being evaluated as unqualified, the wavelet neural network after training is updated according to assessment data, and return according to
The listed average price and conclusion of the business average price of each historical time section of the Target cell source of houses are trained initial wavelet neural network
The step of；When being evaluated as qualification, using the initial wavelet neural network after training as default wavelet neural network.
In one of the embodiments, it is described by the test set to the initial wavelet neural network after the training into
Row assessment obtains assessment result and specifically includes:
By the initial wavelet neural network after test set input training, appraisal test result is obtained；
Initial market average price gap corresponding with the appraisal test result in the appraisal test result is obtained to be no more than
The ratio of the appraisal test result of preset range；
When the ratio is no more than preset threshold, it is evaluated as qualification, when the ratio is more than preset threshold, is evaluated as
It is unqualified.
The half parameter quantile regression is specifically as follows in one of the embodiments:
Y=X β+g (T)+ε
Wherein Y is target source of houses estimated price, and X is the factor for influencing the appraisal of the target source of houses, the source of houses including Target cell
Argument section in the appraisal influence factor of average price and the target source of houses, β are regression coefficient, and g (T) is in appraisal influence factor
Nonparametric part, ε are random error.
The listed average price for obtaining each preset time period of the Target cell source of houses, conclusion of the business in one of the embodiments,
Before average price and initial market average price, further includes:
Classified according to the property type of the Target cell source of houses to the Target cell source of houses；
Listed average price, conclusion of the business average price and the Target cell for obtaining each preset time period of the Target cell source of houses
The initial market average price of the cell source of houses specifically include:
Listed average price, conclusion of the business average price and the source of houses of each preset time period of all kinds of sources of houses of Target cell are obtained respectively
Initial market average price；
It is described that the listed average price, the conclusion of the business average price and the price associated data are inputted into preset wavelet neural
Network, the prediction average value parameter for obtaining the Target cell source of houses specifically include:
Respectively by the listed average price of all kinds of sources of houses of Target cell, the conclusion of the business average price and the price associated data
Preset wavelet neural network is inputted, the prediction average value parameter of all kinds of sources of houses of the Target cell is obtained.
A kind of cell source of houses value parameter estimation device, described device include:
Price data obtains module, for obtaining the listed average price of each preset time period of the Target cell source of houses, striking a bargain
Valence and initial market average price；
Interaction prediction module, for the listed average price, the conclusion of the business average price and the initial market of the cell source of houses
Average price carries out the prediction of the grey multivariable degree of association, obtains price associated data；
Average price prediction module, for inputting the listed average price, the conclusion of the business average price and the price associated data
Preset wavelet neural network obtains the prediction average value parameter of the Target cell source of houses, the preset wavelet neural
Network is obtained using the historical price data of each historical time section of the Target cell source of houses as training data, the historical price number
According to including listed average price, conclusion of the business average price, price associated data and initial market average price；
One room monovalence estimation module, for obtaining each appraisal influence factor of the target source of houses, according to the Target cell room
Each appraisal influence factor of the prediction average value parameter in source and the target source of houses is by presetting half parameter quantile regression mould
Type estimates the value parameter of the target source of houses, and the default half parameter quantile regression is based on each appraisal influence factor and institute
Influence of the prediction average value parameter to the value parameter of the target source of houses is stated to establish.
A kind of computer equipment, including memory and processor, the memory are stored with computer program, the processing
Device performs the steps of when executing the computer program
Obtain listed average price, conclusion of the business average price and the initial market average price of each preset time period of the Target cell source of houses；
The association of grey multivariable is carried out to the listed average price, the conclusion of the business average price and cell source of houses market average price
Degree prediction, obtains price associated data；
The listed average price, the conclusion of the business average price and the price associated data are inputted into preset Wavelet Neural Network
Network obtains the prediction average value parameter of the Target cell source of houses, and the preset wavelet neural network is with Target cell room
The listed average price and conclusion of the business average price of each historical time section in source are obtained as training data；
Obtain the target source of houses each appraisal influence factor, according to the prediction average value parameter of the Target cell source of houses with
And each appraisal influence factor of the target source of houses estimates that the value of the target source of houses is joined by presetting half parameter quantile regression
Number, the default half parameter quantile regression are based on each appraisal influence factor and the prediction average value parameter to institute
The influence for stating the value parameter of the target source of houses is established.
A kind of computer readable storage medium, is stored thereon with computer program, and the computer program is held by processor
It is performed the steps of when row
Obtain listed average price, conclusion of the business average price and the initial market average price of each preset time period of the Target cell source of houses；
The association of grey multivariable is carried out to the listed average price, the conclusion of the business average price and cell source of houses market average price
Degree prediction, obtains price associated data；
The listed average price, the conclusion of the business average price and the price associated data are inputted into preset Wavelet Neural Network
Network obtains the prediction average value parameter of the Target cell source of houses, and the preset wavelet neural network is with Target cell room
The listed average price and conclusion of the business average price of each historical time section in source are obtained as training data；
Obtain the target source of houses each appraisal influence factor, according to the prediction average value parameter of the Target cell source of houses with
And each appraisal influence factor of the target source of houses estimates that the value of the target source of houses is joined by presetting half parameter quantile regression
Number, the default half parameter quantile regression are based on each appraisal influence factor and the prediction average value parameter to institute
The influence for stating the value parameter of the target source of houses is established.
Above-mentioned cell source of houses value parameter estimation method, device, computer equipment and storage medium obtain be used for first
Estimate the basic data of Target cell source of houses average price, i.e. the listed average price of each preset time period of the Target cell source of houses, conclusion of the business is equal
The cell source of houses market average price of valence and Target cell, then to listed average price, conclusion of the business average price and cell source of houses market average price
The prediction of the grey multivariable degree of association is carried out, incidence relation, that is, price associated data between three is obtained；Then by listed average price,
Conclusion of the business average price and price associated data input preset wavelet neural network, estimate to the average price of the Target cell source of houses,
The appraisal parameter for then obtaining the target source of houses, joins according to the appraisal of the average value parameter of the Target cell source of houses and the target source of houses
Number obtains the assessed value parameter of the target source of houses by default half parameter quantile regression.The application, which can use, has transaction
The conclusion of the business of the cell source of houses of completion or listed data establish the valuation model of Target cell, then evaluate mould by Target cell
Type carries out accurate valuation to the average price of Target cell, then by presetting half parameter quantile regression to the price of the target source of houses
Estimated, the transaction data woth no need to the source of houses of house type identical as the target source of houses can carry out the value parameter of the target source of houses
Estimation.
Detailed description of the invention
Fig. 1 is the flow diagram of cell source of houses value parameter estimation method in one embodiment；
Fig. 2 is the flow diagram of cell source of houses value parameter estimation method in one embodiment；
Fig. 3 is the flow diagram of cell source of houses value parameter estimation method in one embodiment；
Fig. 4 is the structural block diagram that cell source of houses value parameter estimates estimation device in one embodiment；
Fig. 5 is the internal structure chart of computer equipment in one embodiment.
Specific embodiment
It is with reference to the accompanying drawings and embodiments, right in order to which the objects, technical solutions and advantages of the application are more clearly understood
The application is further elaborated.It should be appreciated that specific embodiment described herein is only used to explain the application, not
For limiting the application.
Cell source of houses average value method for parameter estimation provided by the present application, estimates for the average price to the cell source of houses
Meter, can specifically be realized, computer journey by cell source of houses value parameter estimation method of the computer program to the application
Sequence can load in terminal, and terminal can be, but not limited to be various personal computers, laptop, smart phone, plate
Computer.
As shown in Figure 1, the cell source of houses average value method for parameter estimation of the application in one of the embodiments, packet
It includes:
S200, the listed average price, conclusion of the business average price and initial market for obtaining each preset time period of the Target cell source of houses are equal
Valence.
Target cell refers to the target of appraisal, can be estimated by average price of this method to the Target cell source of houses
Meter.Each preset time period specifically refers to the multi-section time of distance so far, such as can be for now for the past of starting point
The every month of half a year, the average price of the cell source of houses maintains to stablize within this half a year, and there is no financial crises etc may
To the biggish event of source of houses price of Target cell.Data of the period according to caused by data to the Target cell source of houses
It is grouped.Sticker price refers to the home price that house owner fixs when mechanism, letting agency is listed.Listed average price is listed
The average value of valence.Knock-down price refers to house owner and actual basis of price of the house buyer at conclusion of the business house.Conclusion of the business average price
Refer to the average value of knock-down price.Initial market average price refers to the cell source of houses average price that market provides, wherein an implementation
In example, the source of houses for the Target cell that initial market average price can be provided according to each letting agency website and mechanism, letting agency
Average price obtains.
The listed average price, conclusion of the business average price and initial market for obtaining each preset time period of the Target cell source of houses first are equal
Valence.
S400 carries out the grey multivariable degree of association to listed average price, conclusion of the business average price and the initial market average price of the cell source of houses
Prediction obtains price associated data；
The degree of association refers to the case where mutually changing between things or factor, the size including variation, direction and speed etc.
Relativity, if things or the situation of factor variation are almost the same, it may be considered that the degree of association between them is larger, it is on the contrary then
The degree of association is smaller.Predict that the degree of association compares the statistical analysis side by recurrence or correlation etc by the grey multivariable degree of association
Data required for method is analyzed are less, lower to data demand.
It is equal that listed average price, conclusion of the business average price and the initial market of the cell source of houses can be easily obtained by grey relational grade analysis
Degree of association data between valence.
Listed average price, conclusion of the business average price and price associated data are inputted preset wavelet neural network, obtain mesh by S600
The prediction average price of the cell source of houses is marked, preset wavelet neural network is with the historical price of each historical time section of the Target cell source of houses
Data are obtained as training data, and historical price data include listed average price, conclusion of the business average price, price associated data and initial city
Field average price.
Wavelet neural network is a kind of network system that wavelet theory is combined to birth with neural network theory, i.e., will
The activation primitive of neural network, such as Sigmoid function, are substituted for wavelet function, and corresponding input layer is to the weight of hidden layer
And activation threshold, it is replaced by the scale contraction-expansion factor and the time-shifting factor of wavelet function.Wavelet neural network can be to avoid
Blindness in the design of the structures such as BP neural network；And secondly wavelet neural network has stronger learning ability, and precision is higher.
And because wavelet theory is full size analysis, not only there are globally optimal solution, also holding local detail optimal solution, it is generally, right
Same learning tasks, wavelet neural network structure is simpler, and faster, precision is higher for convergence rate.Preset Wavelet Neural Network
Network, which refers to, has been completed trained neural network model, specifically can be by the listed of each historical time section of the Target cell source of houses
Average price and conclusion of the business average price obtain preset wavelet neural as training data to be trained to initial wavelet neural network
Network.Wavelet neural network has specifically included input layer, hidden layer and output layer.Average value parameter refers to for describing cell
The design parameter of source of houses average value can embody the specific value of the cell source of houses by average value parameter.
Listed average price, conclusion of the business average price and price associated data are inputted into default wavelet neural network, by hidden layer to upper
It states data to be handled, the prediction average value parameter for obtaining the Target cell source of houses is then exported via output layer.
S800 obtains each appraisal influence factor of the target source of houses, according to the prediction average value parameter of the Target cell source of houses
And each appraisal influence factor of the target source of houses, estimate that the value of the target source of houses is joined by presetting half parameter quantile regression
Number presets value of the half parameter quantile regression based on each appraisal influence factor and the prediction average value parameters on target source of houses
The influence of parameter is established.
The appraisal parameter of the target source of houses specifically refer to may include building information where the target source of houses, the target source of houses room
Room position, layout structure, daylighting and area etc. factors.Half parameter quantile regression is specifically as follows:
Y=X β+g (T)+ε
Wherein Y is target source of houses estimated price, and X is the factor for influencing the appraisal of the target source of houses, the source of houses including Target cell
Argument section in the appraisal influence factor of average price and the target source of houses, β are regression coefficient, and g (T) is in appraisal influence factor
Nonparametric part, ε are random error.
Obtain the Target cell source of houses prediction average value parameter after, obtain the target source of houses each appraisal influence because
Element, and the value parameter of the target source of houses is calculated by presetting half parameter quantile regression.Compared with traditional mean regression,
Half parameter quantile regression can depict the situation under the conditions of given independent variable on each quantile of response variable comprehensively.Due to room
The mean regression of source data is unsatisfactory for the normal distribution of error term, and quantile regression is not limited by error term distribution, can be more
Accurately predict as a result, and have explanatory well, carried out so as to accurately value parameter to the target source of houses
Estimation.
In one of the embodiments, before the appraisal affecting parameters for obtaining the target source of houses, we can be according to influence
The condition of the value parameter of the target source of houses, i.e. influence of each variable to source of houses value parameter in half parameter quantile regression,
Establish half parameter quantile regression.Each variable specifically includes the second level factor under the level-one factor and the level-one factor, wherein
In one embodiment, the level-one factor specifically includes the position Lou Dong, construction quality, building capacity, building corollary equipment, house position
It sets, the Multiple factors such as layout structure, daylighting situation and floor space, and the position Lou Dong under the level-one factor includes traffic
The second levels factor such as convenience and ornamental value includes the second levels factors such as purposes, function, exterior wall and service life, building under construction quality
It include that longitudinal capacity, lateral capacity, overall carrying population, structural bearing population, structure universality, overall space are general under capacity
Adaptive and exterior space etc..Also have under building corollary equipment, house location etc. level-one factor corresponding each second level because
Son.Foundation is specifically included by the process of half parameter quantile regression: each variable is determined, then by kernel function according to above-mentioned
Each variable carrys out double of parameter quantile regression and is estimated, can be marked in one of the embodiments, by nonparametric AIC
Standard selects suitable bandwidth, then selects general gaussian kernel function, passes through double of parameter point in Local Polynomial linear approach
Position regression model is estimated, available half parameter quantile regression is obtained.
Above-mentioned cell source of houses value parameter estimation method obtains the basic number for estimating Target cell source of houses average price first
According to, i.e., the listed average price of each preset time period of the Target cell source of houses in preset time period, conclusion of the business average price and Target cell
Cell source of houses market average price then carries out grey multivariable pass to listed average price, conclusion of the business average price and cell source of houses market average price
The prediction of connection degree, obtains incidence relation, that is, price associated data between three；Then by listed average price, conclusion of the business average price and price
Associated data inputs preset wavelet neural network, estimates to the average value parameter of the Target cell source of houses.The application can
To be predicted using the conclusion of the business or listed data correlation degree that have the cell source of houses that transaction is completed, then by listed average price, at
Average price and degree of association prediction is handed over accurately to be estimated according to average value parameter of the preset wavelet neural network to Target cell
Value, it is more objective and accurate compared to the subjectivity appraisal of appraiser.
As shown in Fig. 2, in one of the embodiments, before step S200, further includes:
S120 obtains the listed price data and concluded price data of each preset time period of the Target cell source of houses.
S140 obtains the listed average price of each preset time period of the Target cell source of houses according to listed price data, according at
Price data is handed over, the conclusion of the business average price of each preset time period of the Target cell source of houses is obtained.
The listed price data and concluded price data of each preset time period of the Target cell source of houses available first,
The listed average price of the target source of houses each period is then calculated according to listed price data and concluded price data and is struck a bargain
Average price, it is simple and convenient.
In one of the embodiments, before step S140 further include:
S130 is filtered processing to listed price data and concluded price data based on mean filter principle.
Filtration treatment refers to reject in listed price data and concluded price data abnormal data occur.Such as it can be with
Based on mean filter principle, exclude initial below or above the cell source of houses at that time in listed price data and concluded price data
The data of market average price 40%.Or the data that listed price and concluded price have big difference.Since the price of the source of houses is easy
It is affected by various factors, for example some is sold room person is eager to spend money, it is possible to sell the cell source of houses with too low price.Again
Alternatively, some source of houses possible position of Target cell is too high, so that the source of houses price is too low.Or multiple house purchasers are simultaneously
It is interested in some source of houses, the mode bidded is then passed through to want to buy the source of houses, so as to be higher than current area equal for the source of houses price
Valence.These abnormal prices are not high for the reference value for calculating cell average price, so using the data as abnormal data from calculating
It is rejected in the prediction average price of Target cell.By the listed price data and knock-down price of rejecting the Target cell source of houses
Abnormal data in lattice data can sufficiently improve the validity of the prediction average price of the resulting Target cell source of houses.
As shown in figure 3, in one of the embodiments, before step S600, further includes:
S510, obtains the historical price data of each historical time section of the Target cell source of houses, and historical price data include
Listed average price, conclusion of the business average price, price associated data and initial market average price, the average price that is listed, conclusion of the business average price and price association
Data are input data, and initial market average price is prediction result.
Historical price data are divided into training set and test set at random by S530；
S550, by training set input initial wavelet neural network, by gradient coaching method to initial wavelet neural network into
Row training；
S570 assesses the initial wavelet neural network after training by test set.
S592 obtains assessment result, and when being evaluated as unqualified, enters step S594, updates training according to assessment data
Wavelet neural network afterwards, and return to the listed average price and conclusion of the business average price of each historical time section according to the Target cell source of houses
The step of initial wavelet neural network is trained；When being evaluated as qualification, S596 is entered step, it will be initial after training
Wavelet neural network is as default wavelet neural network.
Wherein step S570 can specifically include:
By the initial wavelet neural network after test set input training, appraisal test result is obtained.
It obtains initial market average price gap corresponding with appraisal test result in appraisal test result and is no more than preset range
Appraisal test result ratio.
When ratio is no more than preset threshold, it is evaluated as qualification, when ratio is more than preset threshold, is evaluated as unqualified.
Data in test set and training set are specifically each historical price data being grouped according to the period, historical price number
According to specifically including listed average price, conclusion of the business average price, price associated data and initial market average price.Average price, conclusion of the business be wherein listed
Valence and price associated data are the parameters of input model, and initial market average price is the target component of model estimation.
Can listed average price based on each period of the history of the Target cell source of houses and conclusion of the business average price to small echo mind
It is trained through network.Such as this month is August, the preset time in this programme refers to the 2-7 month, can by the method for this programme
To be predicted based on the Target cell 2 months listed prices and concluded price to July the source of houses average price of Target cell, obtain
Obtain the prediction average price of Target cell.What the data wherein inputted included is 2 months listed average price, the conclusion of the business average prices to July, defeated
The prediction average price of acquisition Target cell out.But by this 6 months Target cell January to June or last Dec to the present
Its May this 6 months listed average price and conclusion of the business average price wavelet neural network is trained as sample.By pervious small
Area's source of houses price is trained wavelet neural network, can effectively improve the accuracy of wavelet neural network prediction.
Assessment refers to that the wavelet neural network completed to training is assessed, and judges whether its accuracy rate reaches expected, i.e.,
By in test set each group historical price data input training of judgement complete wavelet neural network, obtain its prediction as a result,
The obtained estimated price of model is obtained to be not less than or account for higher than 10% of the initial market average price in the historical price data
The ratio of all test datas judges whether the ratio is more than preset threshold value, i.e. the estimated target obtained of judgment models is small
Whether the error of the source of houses average value parameter in area is in preset acceptable range, when it is in preset error range
When, judge that its training is completed, can be used for estimating the source of houses average price of Target cell.
It further include the property type according to the Target cell source of houses to mesh in one of the embodiments, before step S200
The mark cell source of houses is classified.
Step S200 is specifically included: respectively obtain each preset time period of all kinds of sources of houses of Target cell listed average price, at
Hand over average price and the initial market average price of the source of houses.
Step S600 is specifically included: respectively closing the listed average price, conclusion of the business average price and price of all kinds of sources of houses of Target cell
Join data and input preset wavelet neural network, obtains the prediction average price of all kinds of sources of houses of Target cell.
It is obtaining in preset time before the listed price of the Target cell source of houses, it can be according to the property type of the source of houses to the source of houses
Classify, source of houses average price is then sought according to the type of each source of houses.The property type of the cell source of houses is varied, such as not
Villa, retail shop, apartment and ordinary commercial housing etc..The type of the source of houses has this tremendous influence to the price of the source of houses, can pass through root
Classify according to the property type of the cell source of houses to the cell source of houses, then calculates separately the flat of the source of houses of every a kind of property type
Equal price improves the validity of the prediction average price of the resulting Target cell source of houses by distinguishing property type, prevents different
Influence of the source of houses of normal type for the cell average price predicted.
The cell source of houses value parameter estimation method of the application in one of the embodiments, comprising: obtain Target cell
The listed price data and concluded price data of each preset time period of the source of houses.To listed price data and concluded price number
It is handled according to being filtered.The listed average price that each preset time period of the Target cell source of houses is obtained according to listed price data, according to
The listed average price of each preset time period of the Target cell source of houses and the conclusion of the business average price of each preset time period of the Target cell source of houses.It obtains
Take the initial market average price of listed average price, conclusion of the business average price and the cell source of houses of each preset time period of the Target cell source of houses.To extension
Board average price, conclusion of the business average price and the initial market average price of the cell source of houses carry out the prediction of the grey multivariable degree of association, obtain price association
Data.Obtain the historical price data of each historical time section of the Target cell source of houses, historical price data include listed average price,
Conclusion of the business average price, price associated data and initial market average price, be listed average price, conclusion of the business average price and price associated data are input
Data, initial market average price are prediction result；Historical price data are divided into training set and test set at random；Training set is inputted
Initial wavelet neural network is trained initial wavelet neural network by gradient coaching method；By test set to training after
Initial wavelet neural network assessed, obtain assessment result；When being evaluated as unqualified, training is updated according to assessment data
Wavelet neural network afterwards, and return to the listed average price and conclusion of the business average price of each historical time section according to the Target cell source of houses
The step of initial wavelet neural network is trained；When being evaluated as qualification, the initial wavelet neural network after training is made
To preset wavelet neural network.Listed average price, conclusion of the business average price and price associated data can then be inputted when assessing qualified pre-
If wavelet neural network, obtain the Target cell source of houses prediction average value parameter.Finally obtain each appraisal of the target source of houses
Influence factor passes through pre- according to each appraisal influence factor of the prediction average value parameter of the Target cell source of houses and the target source of houses
If the value parameter of the half parameter quantile regression estimation target source of houses.
The cell source of houses value parameter estimation method of the application passes through computer software reality in one of the embodiments,
Existing, user wishes to inquire the price of some source of houses in some cell, prepares to buy house.When he is to including the application cell
The computer software input of source of houses value parameter estimation method wishes after inquiring the title of the cell of average price that the software obtains first
Listed average price, conclusion of the business average price and the cell in the source of houses each month of Target cell (i.e. the cell of user's input) in half a year in past
The initial market average price of the source of houses；It is changeable that grey then is carried out to listed average price, conclusion of the business average price and the initial market average price of the cell source of houses
Degree of association prediction is measured, price associated data is obtained；Finally listed average price, conclusion of the business average price and price associated data are inputted default
Wavelet neural network, obtain the Target cell source of houses prediction average price.Finally obtain each appraisal influence factor of the target source of houses, root
Pass through default half parameter point according to the prediction average value parameter of the Target cell source of houses and each appraisal influence factor of the target source of houses
The value parameter of the position regression model estimation target source of houses.The predictive value parameter of the target source of houses is then shown to user, for
Family reference.
It should be understood that although each step in the flow chart of Fig. 1-3 is successively shown according to the instruction of arrow,
These steps are not that the inevitable sequence according to arrow instruction successively executes.Unless expressly stating otherwise herein, these steps
Execution there is no stringent sequences to limit, these steps can execute in other order.Moreover, at least one in Fig. 1-3
Part steps may include that perhaps these sub-steps of multiple stages or stage are not necessarily in synchronization to multiple sub-steps
Completion is executed, but can be executed at different times, the execution sequence in these sub-steps or stage is also not necessarily successively
It carries out, but can be at least part of the sub-step or stage of other steps or other steps in turn or alternately
It executes.
As shown in figure 4, the application also provides a kind of cell source of houses value parameter estimation device, device includes:
Price data obtains module 200, for obtaining listed average price, the conclusion of the business of each preset time period of the Target cell source of houses
Average price and initial market average price；
Interaction prediction module 400, for carrying out ash to listed average price, conclusion of the business average price and the initial market average price of the cell source of houses
The prediction of the color multivariable degree of association, obtains price associated data；
Average price prediction module 600, for listed average price, conclusion of the business average price and price associated data to be inputted preset small echo
Neural network obtains the prediction average value parameter of the Target cell source of houses, and preset wavelet neural network is with the Target cell source of houses
The historical price data of each historical time section are obtained as training data, and historical price data include listed average price, strike a bargain
Valence, price associated data and initial market average price；
One room monovalence estimation module 800, for obtaining each appraisal influence factor of the target source of houses, according to the Target cell source of houses
Prediction average value parameter and the target source of houses each appraisal influence factor, pass through default half parameter quantile regression estimation
The value parameter of the target source of houses presets half parameter quantile regression and is based on each appraisal influence factor and prediction average value parameter
Influence to the value parameter of the target source of houses is established.
In one of the embodiments, further include average price computing module, for obtain the Target cell source of houses it is each default when
Between section listed price data and concluded price data；According to listed price data, it is each default to obtain the Target cell source of houses
The listed average price of period obtains the conclusion of the business average price of each preset time period of the Target cell source of houses according to concluded price data.
It in one of the embodiments, further include data filtering module, for being based on mean filter principle to listed price
Data and concluded price data are filtered processing.
It in one of the embodiments, further include model training module, for obtaining each history of the Target cell source of houses
The historical price data of period, historical price data include listed average price, conclusion of the business average price, price associated data and initial city
Field average price, be listed average price, conclusion of the business average price and price associated data are input data, and initial market average price is prediction result；With
Historical price data are divided into training set and test set by machine；Training set is inputted into initial wavelet neural network, passes through gradient training
Method is trained initial wavelet neural network；The initial wavelet neural network after training is assessed by test set, is obtained
Obtain assessment result；When being evaluated as unqualified, the wavelet neural network after training is updated according to assessment data, and return according to mesh
What the listed average price and conclusion of the business average price for marking each historical time section of the cell source of houses were trained initial wavelet neural network
Step；When being evaluated as qualification, using the initial wavelet neural network after training as default wavelet neural network.
Model training module is also used to inputting test set into the initial wavelet nerve after training in one of the embodiments,
Network obtains appraisal test result；Obtain initial market average price gap corresponding with appraisal test result in appraisal test result
No more than the ratio of the appraisal test result of preset range；When ratio is no more than preset threshold, it is evaluated as qualification, when ratio is super
When crossing preset threshold, it is evaluated as unqualified.
It in one of the embodiments, further include source of houses categorization module, for the property type according to the Target cell source of houses
Classify to the Target cell source of houses.It is also each pre- with all kinds of sources of houses of Target cell are obtained respectively that price data obtains module 200
If the listed average price of period, conclusion of the business average price and the initial market average price of the source of houses.Average price prediction module 600 is also used to mesh respectively
The listed average price, conclusion of the business average price and price associated data for marking all kinds of sources of houses of cell input preset wavelet neural network, obtain
The prediction average value parameter of all kinds of sources of houses of Target cell.
Specific restriction about cell source of houses value parameter estimation device may refer to be worth above for the cell source of houses
The restriction of method for parameter estimation, details are not described herein.Modules in above-mentioned cell source of houses average value parameter estimation apparatus
It can be realized fully or partially through software, hardware and combinations thereof.Above-mentioned each module can be embedded in the form of hardware or independently of
In processor in computer equipment, it can also be stored in a software form in the memory in computer equipment, in order to locate
It manages device and calls the corresponding operation of the above modules of execution.
In one embodiment, a kind of computer equipment is provided, which can be terminal, internal structure
Figure can be as shown in Figure 5.The computer equipment includes processor, the memory, network interface, display connected by system bus
Screen and input unit.Wherein, the processor of the computer equipment is for providing calculating and control ability.The computer equipment is deposited
Reservoir includes non-volatile memory medium, built-in storage.The non-volatile memory medium is stored with operating system and computer journey
Sequence.The built-in storage provides environment for the operation of operating system and computer program in non-volatile memory medium.The calculating
The network interface of machine equipment is used to communicate with external terminal by network connection.When the computer program is executed by processor with
Realize a kind of cell source of houses value parameter estimation method.The display screen of the computer equipment can be liquid crystal display or electronics
Ink display screen, the input unit of the computer equipment can be the touch layer covered on display screen, are also possible to computer and set
Key, trace ball or the Trackpad being arranged on standby shell, can also be external keyboard, Trackpad or mouse etc..
It will be understood by those skilled in the art that structure shown in Fig. 5, only part relevant to application scheme is tied
The block diagram of structure does not constitute the restriction for the computer equipment being applied thereon to application scheme, specific computer equipment
It may include perhaps combining certain components or with different component layouts than more or fewer components as shown in the figure.
In one embodiment, a kind of computer equipment, including memory and processor are provided, is stored in memory
Computer program, the processor perform the steps of when executing computer program
Obtain listed average price, conclusion of the business average price and the initial market average price of each preset time period of the Target cell source of houses；
The prediction of the grey multivariable degree of association is carried out to listed average price, conclusion of the business average price and the initial market average price of the cell source of houses,
Obtain price associated data；
Listed average price, conclusion of the business average price and price associated data are inputted into preset wavelet neural network, it is small to obtain target
The prediction average value parameter of area's source of houses, preset wavelet neural network is with the history of each historical time section of the Target cell source of houses
Price data is obtained as training data, and historical price data include listed average price, conclusion of the business average price, price associated data and just
Beginning market average price；
Each appraisal influence factor for obtaining the target source of houses, according to the prediction average value parameter and mesh of the Target cell source of houses
Each appraisal influence factor for marking the source of houses estimates the value parameter of the target source of houses by presetting half parameter quantile regression, presets
Value parameter of the half parameter quantile regression based on each appraisal influence factor and the prediction average value parameters on target source of houses
It influences to establish.
In one embodiment, it is also performed the steps of when processor executes computer program and obtains the Target cell source of houses
The listed price data and concluded price data of each preset time period；According to listed price data, Target cell room is obtained
The listed average price of each preset time period in source obtains each preset time period of the Target cell source of houses according to concluded price data
Conclusion of the business average price.
In one embodiment, it also performs the steps of when processor executes computer program based on mean filter principle
Processing is filtered to listed price data and concluded price data.
In one embodiment, it is also performed the steps of when processor executes computer program and obtains the Target cell source of houses
Each historical time section historical price data, historical price data include listed average price, conclusion of the business average price, price associated data
And initial market average price, be listed average price, conclusion of the business average price and price associated data are input data, and initial market average price is pre-
Survey result；Historical price data are divided into training set and test set at random；Training set is inputted into initial wavelet neural network, is passed through
Gradient coaching method is trained initial wavelet neural network；The initial wavelet neural network after training is carried out by test set
Assessment obtains assessment result；When being evaluated as unqualified, the wavelet neural network after training is updated according to assessment data, and return
Return according to the listed average price of each historical time section of the Target cell source of houses and conclusion of the business average price to initial wavelet neural network into
The step of row training；When being evaluated as qualification, using the initial wavelet neural network after training as default wavelet neural network.
In one embodiment, it also performs the steps of to input test set when processor executes computer program and train
Initial wavelet neural network afterwards obtains appraisal test result；It obtains corresponding with appraisal test result in appraisal test result
Initial market average price gap is no more than the ratio of the appraisal test result of preset range；When ratio is no more than preset threshold, comment
Estimate and is evaluated as unqualified for qualification when ratio is more than preset threshold.
In one embodiment, it also performs the steps of when processor executes computer program according to the Target cell source of houses
Property type classify to the Target cell source of houses；The listed of each preset time period of all kinds of sources of houses of Target cell is obtained respectively
Average price, conclusion of the business average price and the initial market average price of the source of houses；Respectively by the listed average price of all kinds of sources of houses of Target cell, conclusion of the business average price with
And price associated data inputs preset wavelet neural network, obtains the prediction average value parameter of all kinds of sources of houses of Target cell.
In one embodiment, a kind of computer readable storage medium is provided, computer program is stored thereon with, is calculated
Machine program performs the steps of when being executed by processor
Obtain listed average price, conclusion of the business average price and the initial market average price of each preset time period of the Target cell source of houses；
The prediction of the grey multivariable degree of association is carried out to listed average price, conclusion of the business average price and the initial market average price of the cell source of houses,
Obtain price associated data；
Listed average price, conclusion of the business average price and price associated data are inputted into preset wavelet neural network, it is small to obtain target
The prediction average value parameter of area's source of houses, preset wavelet neural network is with the history of each historical time section of the Target cell source of houses
Price data is obtained as training data, and historical price data include listed average price, conclusion of the business average price, price associated data and just
Beginning market average price；
Each appraisal influence factor for obtaining the target source of houses, according to the prediction average value parameter and mesh of the Target cell source of houses
Each appraisal influence factor for marking the source of houses estimates the value parameter of the target source of houses by presetting half parameter quantile regression, presets half
Shadow of the parameter quantile regression based on each appraisal influence factor and the value parameter for predicting the average value parameters on target source of houses
It rings and establishes.In one embodiment, it is also performed the steps of when computer program is executed by processor and obtains the Target cell source of houses
The listed price data and concluded price data of each preset time period；According to listed price data, Target cell room is obtained
The listed average price of each preset time period in source obtains each preset time period of the Target cell source of houses according to concluded price data
Conclusion of the business average price.
In one embodiment, it is also performed the steps of when computer program is executed by processor based on mean filter original
Reason is filtered processing to listed price data and concluded price data.
In one embodiment, it is also performed the steps of when processor executes computer program and obtains the Target cell source of houses
Each historical time section historical price data, historical price data include listed average price, conclusion of the business average price, price associated data
And initial market average price, be listed average price, conclusion of the business average price and price associated data are input data, and initial market average price is pre-
Survey result；Historical price data are divided into training set and test set at random；Training set is inputted into initial wavelet neural network, is passed through
Gradient coaching method is trained initial wavelet neural network；The initial wavelet neural network after training is carried out by test set
Assessment obtains assessment result；When being evaluated as unqualified, the wavelet neural network after training is updated according to assessment data, and return
Return according to the listed average price of each historical time section of the Target cell source of houses and conclusion of the business average price to initial wavelet neural network into
The step of row training；When being evaluated as qualification, using the initial wavelet neural network after training as default wavelet neural network.
In one embodiment, it also performs the steps of to input test set when processor executes computer program and train
Initial wavelet neural network afterwards obtains appraisal test result；It obtains corresponding with appraisal test result in appraisal test result
Initial market average price gap is no more than the ratio of the appraisal test result of preset range；When ratio is no more than preset threshold, comment
Estimate and is evaluated as unqualified for qualification when ratio is more than preset threshold.
In one embodiment, it also performs the steps of when computer program is executed by processor according to Target cell room
The property type in source classifies to the Target cell source of houses；The extension of each preset time period of all kinds of sources of houses of Target cell is obtained respectively
Board average price, conclusion of the business average price and the initial market average price of the source of houses；Respectively by the listed average price of all kinds of sources of houses of Target cell, conclusion of the business average price
And price associated data inputs preset wavelet neural network, obtains the prediction average value ginseng of all kinds of sources of houses of Target cell
Number.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Instruct relevant hardware to complete by computer program, computer program to can be stored in a non-volatile computer readable
It takes in storage medium, the computer program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein, this Shen
Please provided by any reference used in each embodiment to memory, storage, database or other media, may each comprise
Non-volatile and/or volatile memory.Nonvolatile memory may include read-only memory (ROM), programming ROM
(PROM), electrically programmable ROM (EPROM), electrically erasable ROM (EEPROM) or flash memory.Volatile memory may include
Random access memory (RAM) or external cache.By way of illustration and not limitation, RAM is available in many forms,
Such as static state RAM (SRAM), dynamic ram (DRAM), synchronous dram (SDRAM), double data rate sdram (DDRSDRAM), enhancing
Type SDRAM (ESDRAM), synchronization link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM
(RDRAM), direct memory bus dynamic ram (DRDRAM) and memory bus dynamic ram (RDRAM) etc..
Each technical characteristic of above embodiments can be combined arbitrarily, for simplicity of description, not to above-described embodiment
In each technical characteristic it is all possible combination be all described, as long as however, the combination of these technical characteristics be not present lance
Shield all should be considered as described in this specification.
Above embodiments only express the several embodiments of the application, and the description thereof is more specific and detailed, but can not
Therefore it is construed as limiting the scope of the patent.It should be pointed out that for those of ordinary skill in the art,
Under the premise of not departing from the application design, various modifications and improvements can be made, these belong to the protection scope of the application.
Therefore, the scope of protection shall be subject to the appended claims for the application patent.
Claims (10)
1. a kind of cell source of houses value parameter estimation method, which comprises
Obtain listed average price, conclusion of the business average price and the initial market average price of each preset time period of the Target cell source of houses；
The prediction of the grey multivariable degree of association is carried out to the listed average price, the conclusion of the business average price and the initial market average price,
Obtain price associated data；
The listed average price, the conclusion of the business average price and the price associated data are inputted into preset wavelet neural network, obtained
The prediction average value parameter of the Target cell source of houses is obtained, the preset wavelet neural network is each with the Target cell source of houses
The historical price data of historical time section are obtained as training data, and the historical price data include listed average price, strike a bargain
Valence, price associated data and initial market average price；
Each appraisal influence factor for obtaining the target source of houses, according to the prediction average value parameter of the Target cell source of houses and institute
Each appraisal influence factor for stating the target source of houses estimates the value parameter of the target source of houses by presetting half parameter quantile regression,
The default half parameter quantile regression is based on each appraisal influence factor with the prediction average value parameter to described
The influence of the value parameter of the target source of houses is established.
2. the method according to claim 1, wherein described obtain the Target cell source of houses each preset time period
Before listed average price, conclusion of the business average price and initial market average price, further includes:
Obtain the listed price data and concluded price data of each preset time period of the Target cell source of houses；
According to the listed price data, obtain the listed average price of each preset time period of the Target cell source of houses, according to it is described at
Price data is handed over, the conclusion of the business average price of each preset time period of the Target cell source of houses is obtained.
3. according to the method described in claim 2, acquisition target is small it is characterized in that, described according to the listed price data
It is each default to obtain the Target cell source of houses according to the concluded price data for the listed average price of each preset time period of area's source of houses
Before the conclusion of the business average price of period, further includes:
Processing is filtered to the listed price data and the concluded price data based on mean filter principle.
4. the method according to claim 1, wherein it is described by the listed average price, the conclusion of the business average price and
The price associated data inputs preset wavelet neural network, obtains the prediction average value parameter of the Target cell source of houses
Before, further includes:
The historical price data of each historical time section of the Target cell source of houses are obtained, the historical price data include hanging
Board average price, conclusion of the business average price, price associated data and initial market average price, the listed average price, conclusion of the business average price and price are closed
Connection data are input data, and the initial market average price is prediction result；
The historical price data are divided into training set and test set at random；
The training set is inputted into initial wavelet neural network, initial wavelet neural network is instructed by gradient coaching method
Practice；
The initial wavelet neural network after the training is assessed by the test set, obtains assessment result；
When being evaluated as unqualified, the wavelet neural network after training is updated according to assessment data, and return according to the target
The step that the listed average price and conclusion of the business average price of each historical time section of the cell source of houses are trained initial wavelet neural network
Suddenly；When being evaluated as qualification, using the initial wavelet neural network after training as default wavelet neural network.
5. according to the method described in claim 4, it is characterized in that, it is described by the test set to initial after the training
Wavelet neural network is assessed, and is obtained assessment result and is specifically included:
By the initial wavelet neural network after test set input training, appraisal test result is obtained；
Obtain in the appraisal test result initial market average price gap corresponding with the appraisal test result be no more than it is default
The ratio of the appraisal test result of range；
When the ratio is no more than preset threshold, it is evaluated as qualification, when the ratio is more than preset threshold, is evaluated as not conforming to
Lattice.
6. the method according to claim 1, wherein the half parameter quantile regression is specifically as follows:
Y=X β+g (T)+ε
Wherein Y is target source of houses estimated price, and X is the factor for influencing the appraisal of the target source of houses, the source of houses average price including Target cell
And the argument section in the appraisal influence factor of the target source of houses, β are regression coefficient, g (T) is the non-ginseng evaluated in influence factor
Number part, ε is random error.
7. the method according to claim 1, wherein described obtain the Target cell source of houses each preset time period
Before listed average price, conclusion of the business average price and initial market average price, further includes:
Classified according to the property type of the Target cell source of houses to the Target cell source of houses；
It is described to obtain the small of the listed average price of each preset time period of the Target cell source of houses, conclusion of the business average price and the Target cell
Source of houses initial market average price in area's specifically includes:
The listed average price, conclusion of the business average price and the source of houses for obtaining each preset time period of all kinds of sources of houses of Target cell respectively are initial
Market average price；
It is described that the listed average price, the conclusion of the business average price and the price associated data are inputted into preset Wavelet Neural Network
Network, the prediction average value parameter for obtaining the Target cell source of houses specifically include:
The listed average price of all kinds of sources of houses of Target cell, the conclusion of the business average price and the price associated data are inputted respectively
Preset wavelet neural network obtains the prediction average value parameter of all kinds of sources of houses of the Target cell.
8. a kind of cell source of houses value parameter estimation device, which is characterized in that described device includes:
Price data obtain module, for obtain listed average price, the conclusion of the business average price of each preset time period of the Target cell source of houses with
And initial market average price；
Interaction prediction module, for the listed average price, the conclusion of the business average price and the initial market average price of the cell source of houses
The prediction of the grey multivariable degree of association is carried out, price associated data is obtained；
Average price prediction module, it is default for inputting the listed average price, the conclusion of the business average price and the price associated data
Wavelet neural network, obtain the prediction average value parameter of the Target cell source of houses, the preset wavelet neural network
It is obtained using the historical price data of each historical time section of the Target cell source of houses as training data, the historical price data packet
Include listed average price, conclusion of the business average price, price associated data and initial market average price；
One room monovalence estimation module, for obtaining each appraisal influence factor of the target source of houses, according to the Target cell source of houses
Each appraisal influence factor of prediction average value parameter and the target source of houses is estimated by default half parameter quantile regression
Count the value parameter of the target source of houses, the default half parameter quantile regression be based on each appraisal influence factor with it is described pre-
Influence of the average value parameter to the value parameter of the target source of houses is surveyed to establish.
9. a kind of computer equipment, including memory and processor, the memory are stored with computer program, feature exists
In the step of processor realizes any one of claims 1 to 7 the method when executing the computer program.
10. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program
The step of method described in any one of claims 1 to 7 is realized when being executed by processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811289888.XA CN109583940A (en) | 2018-10-31 | 2018-10-31 | Cell source of houses value parameter estimation method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811289888.XA CN109583940A (en) | 2018-10-31 | 2018-10-31 | Cell source of houses value parameter estimation method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109583940A true CN109583940A (en) | 2019-04-05 |
Family
ID=65921049
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811289888.XA Pending CN109583940A (en) | 2018-10-31 | 2018-10-31 | Cell source of houses value parameter estimation method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109583940A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110119893A (en) * | 2019-05-08 | 2019-08-13 | 重庆斐耐科技有限公司 | A kind of house property appraisal procedure and system based on big data technology |
| CN110634014A (en) * | 2019-07-19 | 2019-12-31 | 北京无限光场科技有限公司 | Method, device, equipment and medium for determining house source price |
| CN113706212A (en) * | 2021-08-31 | 2021-11-26 | 深圳壹账通智能科技有限公司 | Prediction model-based housing valuation method, device, equipment and storage medium |
-
2018
- 2018-10-31 CN CN201811289888.XA patent/CN109583940A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110119893A (en) * | 2019-05-08 | 2019-08-13 | 重庆斐耐科技有限公司 | A kind of house property appraisal procedure and system based on big data technology |
| CN110634014A (en) * | 2019-07-19 | 2019-12-31 | 北京无限光场科技有限公司 | Method, device, equipment and medium for determining house source price |
| CN113706212A (en) * | 2021-08-31 | 2021-11-26 | 深圳壹账通智能科技有限公司 | Prediction model-based housing valuation method, device, equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Hong et al. | Probabilistic electric load forecasting: A tutorial review | |
| King et al. | Big prisons, small towns: Prison economics in rural America | |
| Chen et al. | Performance risk assessment in public–private partnership projects based on adaptive fuzzy cognitive map | |
| Wilson et al. | Company failure prediction with limited information: Newly incorporated companies | |
| CN109583940A (en) | Cell source of houses value parameter estimation method and device | |
| Amri et al. | Performance of multiple linear regression and nonlinear neural networks and fuzzy logic techniques in modelling house prices | |
| CN104995644A (en) | Institution simulation | |
| Niu et al. | House price expectations | |
| Glumac et al. | Real Estate and Land Property Automated Valuation Systems: A Taxonomy and Conceptual Model. | |
| Rogers et al. | The inclusion of non-market values in systematic conservation planning to enhance policy relevance | |
| Jagtiani et al. | Predicting inadequate capitalization: Early warning system for bank supervision | |
| CN112686470A (en) | Power grid saturation load prediction method and device and terminal equipment | |
| JP2022007877A (en) | Real Estate Transaction Price Proposal Program | |
| Zeng | [Retracted] Research on Risk Measurement and Early Warning of Electronic Banking Business Based on GMDH Algorithm | |
| CN109583939A (en) | Cell source of houses average value method for parameter estimation and device | |
| CN109544410A (en) | Cell source of houses value parameter estimation method and device | |
| Pancrazi et al. | Natural expectations and home equity extraction | |
| Amissah et al. | Is Religion a Determinant of Financial Development? | |
| Demetriou | GIS-based automated valuation models (AVMs) for land consolidation schemes | |
| Yilmaz | Housing GANs: Deep Generation of Housing Market Data | |
| Larrain | The PMI, the T‐Bill and Inventories: A Comparative Analysis of Neural Network and Regression Forecasts | |
| Vernon-Bido et al. | The effect of neighborhood density and GIS layout on the foreclosure contagion effect | |
| Sun et al. | [Retracted] Wheat Futures Prices Prediction in China: A Hybrid Approach | |
| Oyedeji Joseph | Property rental value classification model: a case of osogbo, osun state, Nigeria | |
| US20230027774A1 (en) | Smart real estate evaluation system |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20190405 |