CN110264229A - Used car pricing method based on full Connection Neural Network, device and system - Google Patents
Used car pricing method based on full Connection Neural Network, device and system Download PDFInfo
- Publication number
- CN110264229A CN110264229A CN201810201233.6A CN201810201233A CN110264229A CN 110264229 A CN110264229 A CN 110264229A CN 201810201233 A CN201810201233 A CN 201810201233A CN 110264229 A CN110264229 A CN 110264229A
- Authority
- CN
- China
- Prior art keywords
- used car
- price
- pricing model
- neural network
- adjustment
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/06—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
- G06N3/061—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons using biological neurons, e.g. biological neurons connected to an integrated circuit
-
- 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
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
- G06Q30/0206—Price or cost determination based on market factors
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Strategic Management (AREA)
- Finance (AREA)
- Development Economics (AREA)
- Accounting & Taxation (AREA)
- Molecular Biology (AREA)
- Data Mining & Analysis (AREA)
- Neurology (AREA)
- Entrepreneurship & Innovation (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Microelectronics & Electronic Packaging (AREA)
- Game Theory and Decision Science (AREA)
- Economics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the present application shows a kind of used car pricing method based on full Connection Neural Network, device and system.Method shown in the embodiment of the present application, vehicle characteristics are constructed into full Connection Neural Network by neuroid, calculated result based on the neuron, generate used car pricing model, in the building process of full Connection Neural Network, method shown in the embodiment of the present application, using the building of multi-layer network, pass through layer-by-layer eigentransformation, character representation by vehicle characteristics in former space transforms to a new feature space, feature after ultimately generating building, feature construction used car pricing model after being then based on building, feature after the building that the embodiment of the present application generates, various dimensions consider the contribution that vehicle characteristics are fixed a price in used car, used car pricing model is generated using the feature after the building shown in the embodiment of the present application, the result error predicted in the price-setting process of used car is small.
Description
Technical field
The present invention relates to field of computer technology, in particular to a kind of used car price side based on full Connection Neural Network
Method, device and system.
Background technique
As the development of economic society and life of urban resident level improve, vehicle has become private basic demand.Closely
The trade deal of Nian Lai, the rapid development of China's economic, the rapid growth of vehicle population, used car are more and more flourishing, the whole nation
The second-hand vehicle broker of each provincial capital and unit generally reach 5000 or more, and the used car to strike a bargain every year reaches 10
Ten thousand -20 ten thousand.With the gradually prosperity of the trade deal of used car, used automobile market faces series of challenges therewith, for example, mesh
Preceding used automobile market is still the market of an information asymmetry, and consumer is difficult to know the value of used car, as a result, two
Handcart market is difficult to obtain the trust of consumer, and used automobile market is caused to lose some potential clients.Therefore how couple two
Handcart carries out assessment price, it appears particularly important.
Present used car assessment price is substantially determined that is, appraiser is according to used car by the experience of appraiser
Some surface appearances and personal experience assess.However, above-mentioned assessment pricing method is due to there is excessive human factor to participate in,
It is not accurate enough to often result in assessment result, to cause trouble to both parties.
The prior art uses learning organization established model, is assessed using the model of building the price of used car.In benefit
During constructing used car pricing model with machine learning method, vehicle characteristics are input to computer and are modeled, then,
The model of building is applied to the price of used car.For example, in building used car pricing model, generally according to historical trading
Data accurately predict the price of used car, remove to construct some assemblage characteristics first against existing data, for example, vehicle it is total in
Journey is divided by vehicle age.The assemblage characteristic of prior art building lacks validity and comprehensive, is constructed using prior art characteristic of human nature
Model, to used car carry out price prediction, deviation is larger between prediction result and the concluded price of practical used car.
Summary of the invention
Goal of the invention of the invention is to provide a kind of used car pricing method based on full Connection Neural Network, device,
And system, to solve the validity and comprehensive that the temporal aspect of prior art building lacks, using the timing shown in the prior art
The model of feature construction carries out the prediction of price, deviation between prediction result and the concluded price of practical used car to used car
Larger technical problem.
The embodiment of the present application first aspect shows a kind of used car pricing method based on full Connection Neural Network, the side
Method includes:
Vehicle characteristics are obtained, construct neuron according to the vehicle characteristics；
According to the neuron, full Connection Neural Network is constructed；
According to the full Connection Neural Network, the neuron is successively calculated, generates used car pricing model；
According to the used car pricing model, the price of target used car is calculated.
It is selectable, the acquisition vehicle characteristics, according to the vehicle characteristics construct neuron the step of specifically:
By the vehicle characteristics, neuron is constructed according to following formula:
Z=a1w1+……akwk+……aKwK+b；
Wherein, α=σ (wTA+b)=σ (w1a1+w2a2+…+wkak+b)；
a1,a2,...akFor vehicle characteristics, w1,w2,...wkFor the weight of vehicle characteristics, b is that bias term is second-hand for adjusting
Vehicle pricing model；σ is that activation primitive is used to vehicle characteristics transforming to a new feature space by the character representation in former space.
Selectable, the full Connection Neural Network of basis successively calculates the neuron, generates used car pricing model
The step of include:
According to the full Connection Neural Network, the neuron is successively calculated, generates the first pricing model；
According to first pricing model, the inspection price of verification used car is calculated；
Judge whether the checkout price and the concluded price for verifying used car are consistent,
If consistent, determine that first pricing model is used car pricing model
If inconsistent, first pricing model is adjusted, generates used car pricing model.
It is selectable, the first pricing model of the adjustment, generate used car pricing model the step of include:
It calculates the checkout price and verifies the difference of the concluded price of used car, adjust the ginseng of first pricing model
Number, obtains intermediate pricing model；
According to the intermediate pricing model, the tune inspection price for verifying second-hand vehicle model is calculated；
Judge that described adjust examines price whether consistent with the concluded price of verification used car；
If inconsistent, continue the parameter for adjusting intermediate pricing model, until adjust examine price and verify used car at
Hand over price consistent.
It is selectable, it is described to continue the step of adjusting the parameter of intermediate pricing model specifically:
Determine the adjustment number of the intermediate pricing model；
The adjustment amplitude of parameter is determined according to the adjustment number；
According to the adjustment amplitude, the parameter of intermediate pricing model is adjusted.
It is selectable, the adjustment amplitude that parameter is determined according to adjustment number specifically:
By exponential attenuation method, adjustment amplitude is determined according to the adjustment number, the adjustment amplitude is with the adjustment
Number increases and reduces.
Selectable, the full Connection Neural Network of basis successively calculates the neuron, generates used car pricing model
The step of include:
According to the full Connection Neural Network, the neuron is successively calculated, modeling data is obtained；
Priori is added in the modeling data, the space of the modeling data is reduced, obtains Target Modeling data；
According to the Target Modeling data, used car pricing model is constructed.
The embodiment of the present application second aspect shows a kind of used car pricing device based on full Connection Neural Network, the dress
It sets and includes:
Acquiring unit constructs neuron according to the vehicle characteristics for obtaining vehicle characteristics；
Construction unit, for constructing full Connection Neural Network according to the neuron；
Generation unit generates used car price for successively calculating the neuron according to the full Connection Neural Network
Model；
Computing unit, for calculating the price of target used car according to the used car pricing model.
Selectable, the construction unit is also used to:
By the vehicle characteristics, neuron is constructed according to following formula:
Z=a1w1+……akwk+……aKwK+b；
Wherein, α=σ (wTA+b)=σ (w1a1+w2a2+…+wkak+b)；
a1,a2,…akFor vehicle characteristics, w1,w2,…wkFor the weight of vehicle characteristics, b is bias term for adjusting used car
Pricing model；σ is that activation primitive is used to vehicle characteristics transforming to a new feature space by the character representation in former space.
Selectable, the generation unit includes:
It is fixed to generate first for successively calculating the neuron according to the full Connection Neural Network for first generation unit
Valence model；
Calculation of price unit is examined, for calculating the inspection price of verification used car according to first pricing model；
First judging unit, for judging whether the checkout price and the concluded price for verifying used car are consistent,
Determination unit, if determining that first pricing model is used car pricing model for consistent
Adjustment unit generates used car pricing model if adjusting first pricing model for inconsistent.
Selectable, the adjustment unit includes:
Intermediate pricing model generation unit, the difference of the concluded price for calculating the checkout price and verification used car
Value adjusts the parameter of first pricing model, obtains intermediate pricing model；
Calculation of price unit is examined, for according to the intermediate pricing model, calculating the tune inspection for verifying second-hand vehicle model
Price；
Second judgment unit, for judging that described adjust examines price whether consistent with the concluded price of verification used car；
Continue adjustment unit, if continuing the parameter for adjusting intermediate pricing model for inconsistent, examines price until adjusting
It is consistent with the verification concluded price of used car.
Selectable, the continuation adjustment unit includes:
Number determination unit is adjusted, for determining the adjustment number of the intermediate pricing model；
Adjustment amplitude determination unit, for determining the adjustment amplitude of parameter according to the adjustment number；
Parameter adjustment unit, for adjusting the parameter of intermediate pricing model according to the adjustment amplitude.
Selectable, the adjustment amplitude determination unit is also used to,
By exponential attenuation method, adjustment amplitude is determined according to the adjustment number, the adjustment amplitude is with the adjustment
Number increases and reduces.
Selectable, the construction unit includes:
Layer-by-layer computing unit obtains modeling number for successively calculating the neuron according to the full Connection Neural Network
According to；
Adding unit reduces the space of the modeling data, obtains target and build for priori to be added in the modeling data
Modulus evidence；
Used car pricing model construction unit, for constructing used car pricing model according to the Target Modeling data.
The application leads the embodiment third aspect and shows a kind of used car pricing system based on full Connection Neural Network, described
System includes: application platform server, and the data storage server being connected with the application platform server, the data are deposited
Storage server is arranged inside the Platform Server or is independently arranged, and the application platform server passes through internet and terminal
It is connected；
The terminal is used for the display of target used car price；
The storage that related data is used for according to storage server；
The application platform server is for realizing the method shown in the embodiment of the present application.
From the above technical scheme, the embodiment of the present application shows a kind of used car price based on full Connection Neural Network
Method, device and system, which comprises obtain vehicle characteristics, construct neuron according to the vehicle characteristics；According to institute
Neuron is stated, full Connection Neural Network is constructed；According to the full Connection Neural Network, the neuron is successively calculated, generates two
Handcart pricing model；According to the used car pricing model, the price of target used car is calculated.Side shown in the embodiment of the present application
Vehicle characteristics are constructed full Connection Neural Network by neuroid by method, based on the calculated result of the neuron, generate two
Handcart pricing model, in the building process of full Connection Neural Network, method shown in the embodiment of the present application, using multi-level net
The building of network, by layer-by-layer eigentransformation, the character representation by vehicle characteristics in former space transforms to a new feature space, most
Throughout one's life at the feature after building, the feature construction used car pricing model being then based on after constructing, what the embodiment of the present application generated
Feature after building, various dimensions consider the contribution that vehicle characteristics are fixed a price in used car, using the building shown in the embodiment of the present application
Feature afterwards generates used car pricing model, and the result error predicted in the price-setting process of used car is small.
Detailed description of the invention
It in order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, below will be to institute in embodiment
Attached drawing to be used is needed to be briefly described, it should be apparent that, the accompanying drawings in the following description is only some implementations of the invention
Example, for those of ordinary skill in the art, without creative efforts, can also obtain according to these attached drawings
Obtain other attached drawings.
Fig. 1-1 is the used car pricing system that the full Connection Neural Network of a kind of basis exemplified is preferably implemented according to one
Structural block diagram；
Fig. 1-2 is the used car pricing system according to the full Connection Neural Network of a kind of basis shown in another preferred embodiment
Structural block diagram；
Fig. 2 is the stream for the used car pricing method that the full Connection Neural Network of a kind of basis exemplified is preferably implemented according to one
Cheng Tu；
Fig. 3 is the detail flowchart that the step S103 exemplified is preferably implemented according to one；
Fig. 4 is the detail flowchart that the step S10315 exemplified is preferably implemented according to one；
Fig. 5 is the detail flowchart that the step S103154 exemplified is preferably implemented according to one；
Fig. 6 is the detail flowchart according to the step S103 shown in another preferred embodiment；
Fig. 7 is the knot for the used car pricing device that the full Connection Neural Network of a kind of basis exemplified is preferably implemented according to one
Structure block diagram；
Fig. 8 is the structural block diagram that the generation unit exemplified is preferably implemented according to one；
Fig. 9 is the structural block diagram that the adjustment unit exemplified is preferably implemented according to one；
Figure 10 is the structural block diagram that the continuation adjustment unit exemplified is preferably implemented according to one；
Figure 11 is the structural block diagram according to the generation unit shown in another preferred embodiment；
A kind of structural block diagram of the server exemplified according to Figure 12 by preferred implementation.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Whole description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.According to
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
Embodiment 1:
The prior art uses learning organization established model, is assessed using the model of building the price of used car.In benefit
During constructing used car pricing model with machine learning method, vehicle characteristics are input to computer and are modeled, then,
The model of building is applied to the price of used car.For example, in building used car pricing model, generally according to historical trading
Data accurately predict the price of used car, remove to construct some assemblage characteristics first against existing data, for example, vehicle it is total in
Journey is divided by vehicle age.The assemblage characteristic of prior art building lacks validity and comprehensive, is constructed using prior art characteristic of human nature
Model, to used car carry out price prediction, deviation is larger between prediction result and the concluded price of practical used car.
To solve the above-mentioned problems, the embodiment of the present application first aspect shows a kind of based on the second-hand of full Connection Neural Network
Vehicle pricing system, specifically, Fig. 1-1 is please referred to, and, Fig. 1-2；
The system comprises: application platform server 31, the data storage being connected with the application platform server 31
Server 32, the setting of data storage server 32 is in 31 inside of Platform Server or is independently arranged, and the application is flat
Platform server logical 31 crosses internet and is connected with terminal 33；
The terminal 33 is used for the display of target used car price；
Terminal 33 shown in the embodiment of the present application is the equipment for being in network outermost in computer network, is mainly used for
Input and the output of processing result of user information etc..Mobile terminal shown in the prior art is such as: mobile phone, PAD are in this Shen
It please be in the protection scope of embodiment.
The storage according to storage server 32, for related data；
The application platform server 31, shows method for realizing the embodiment of the present application.
Application platform server 31 shown in the embodiment of the present application is provided a kind of simple and can be managed for web application
The access mechanism to system resource.Application platform server 31 also provides rudimentary service, such as the realization sum number of http protocol
According to library connection management.Servlet container is only a part of application server.Other than Servlet container, application platform clothes
Business device 31 it is also possible to provide other Java EE (Enterprise Edition) component, such as Enterprise Java Bean container, JNDI server with
And JMS service device etc..
Specifically, the application platform server 31, is used for；
(1) vehicle characteristics are obtained, construct neuron according to the vehicle characteristics；
The vehicle characteristics of 1024 used cars of collection of application platform server 31 sample trained as one, then with two
Handcart is unit by the vehicle characteristics input computer of a vehicle, constructs neuron.
(2) according to the neuron, full Connection Neural Network is constructed；
(3) according to the full Connection Neural Network, the neuron is successively calculated, generates used car pricing model；
The full Connection Neural Network includes multilayer neural networks, and every layer of neuroid includes multiple nodes, each
Node corresponds to a neuron；
By taking a used car as an example, building first layer neural network, application platform server 31 collect 1024 two first
The vehicle characteristics of the handcart sample trained as one.Then the vehicle characteristics input of a vehicle is counted as unit of used car
Calculation machine constructs neuron.
Neuron is constructed according to formula below:
Z=a1w1+……akwk+……aKwK+b；
Wherein, α=σ (wTA+b)=σ (w1a1+w2a2+…+wkak+b)；
a1,a2,...akFor vehicle characteristics, w1,w2,...wkFor the weight of vehicle characteristics, b is that bias term is second-hand for adjusting
Vehicle pricing model；σ is that activation primitive is used to vehicle characteristics transforming to a new feature space by the character representation in former space.
Then, using the calculated result of first layer as input value, the second layer is inputted.
Successively transmitting in this way, to the last one layer, the calculated result of final output the last layer, the feature as constructed,
Then according to the feature of building, used car pricing model is constructed.
(4) according to the used car pricing model, the price of target used car is calculated.
System shown in the embodiment of the present application, application platform server use the building of multi-layer network, by successively special
Sign transformation, the character representation by vehicle characteristics in former space transform to a new feature space, the feature after ultimately generating building,
Feature construction used car pricing model after being then based on building, the feature after the building that the embodiment of the present application generates, various dimensions
Consider the contribution that vehicle characteristics are fixed a price in used car, it is fixed to generate used car using the feature after the building shown in the embodiment of the present application
Valence model, the result error predicted in the price-setting process of used car are small.
Embodiment 2:
The embodiment of the present application second aspect shows a kind of used car pricing method based on full Connection Neural Network, specifically
, referring to Fig. 2, the described method includes:
S101 obtains vehicle characteristics, constructs neuron according to the vehicle characteristics；
The vehicle characteristics of the 1024 used cars sample trained as one is collected, then by one as unit of used car
The vehicle characteristics of vehicle input computer, construct neuron.
S102 constructs full Connection Neural Network according to the neuron；
The full Connection Neural Network packet multilayer neural networks, every layer of neuroid include multiple nodes, Mei Gejie
The corresponding neuron of point.
S103 successively calculates the neuron according to the full Connection Neural Network, generates used car pricing model；
The vehicle characteristics of one used car are inputted and are calculated, computer calculates one as a result, then by the by formula
One layer of calculated result inputs second layer neuroid, successively transmits as input value, until the knot of the last layer output
Fruit, the feature as constructed.
S104 calculates the price of target used car according to the used car pricing model.
It can be seen that the building process of the feature shown in the embodiment of the present application, during layer-by-layer calculate, by original vehicle
Feature is mixed, and various dimensions consider the contribution that vehicle characteristics are fixed a price in used car, and the feature after the building is comprehensively true
The price for reacting used car generates used car pricing model using the feature after the building shown in the embodiment of the present application, second-hand
The result error predicted in the price-setting process of vehicle is small.
Embodiment 3:
In order to further increase the embodiment of the present application building feature it is comprehensive, the embodiment of the present application shows a kind of nerve
The construction method of member:
The acquisition vehicle characteristics, according to the vehicle characteristics construct neuron the step of specifically:
By the vehicle characteristics, neuron is constructed according to following formula:
Z=a1w1+……akwk+……aKwK+b；
Wherein, α=σ (wTA+b)=σ (w1a1+w2a2+…+wkak+b)；
a1,a2,...akFor vehicle characteristics, w1,w2,...wkFor the weight of vehicle characteristics, b is that bias term is second-hand for adjusting
Vehicle pricing model；σ is that activation primitive is used to vehicle characteristics transforming to a new feature space by the character representation in former space.
The embodiment of the present application shows method in the building process of full Connection Neural Network, the full Connection Neural Network packet
Multilayer neural networks, every layer of neuroid are configured with corresponding calculation formula；
It is activation primitive that σ is introduced in technology formula shown in the embodiment of the present application, by vehicle characteristics by the feature in former space
Expression transforms to a new feature space.
For example:
The linear activation primitive of the use of the neuron of first layer, the meter for the result input second layer that first layer is calculated
The formula second layer is calculated using hyperbola calculation formula, then the influence by original vehicle characteristics to price is by linear space
It is converted into the hyperbolic space.
It is worth noting that, the embodiment of the present application, which shows technical solution, only illustratively describes vehicle characteristics and valence
The spatial relationship of lattice, in practical applications, the calculation formula of all space conversions that vehicle characteristics and price may be implemented exists
In the protection scope of the embodiment of the present application, here, just not introduced one by one since length is limited.
During feature after the building that the embodiment of the present application generates, various dimensions consider that vehicle characteristics are fixed a price in used car
Contribution, used car pricing model is generated using the feature after the building shown in the embodiment of the present application, in the price of used car
The result error predicted in journey is small.
Embodiment 4:
In order to further increase the accuracy that the embodiment of the present application shows the used car pricing model of method building, the application
The building process for implementing the full Connection Neural Network exemplified includes propagated forward and backpropagation.
Specifically, please referring to Fig. 3:
Embodiment 4 has similar step to the technical solution shown in embodiment 3, only difference is that embodiment 3 is shown
Technical scheme steps S103 out includes the following steps:
S10311 successively calculates the neuron according to the full Connection Neural Network, generates the first pricing model；
S10312 calculates the inspection price of verification used car according to first pricing model；
The verification used car is the used car to have struck a bargain, and is accurately known, the rolling stock feature of used car is verified,
And the concluded price of the verification used car.
The vehicle characteristics of above-mentioned verification used car are inputted into the first pricing model, first pricing model is based on the school
The vehicle characteristics for testing used car calculate a checkout price；
S10313 judges whether the checkout price and the concluded price for verifying used car are consistent；
If consistent, S10314 determines that first pricing model is used car pricing model；
If inconsistent, S10315 adjusts first pricing model, generates used car pricing model.
If the concluded price of checkout price and verification used car is inconsistent, then is fixed a price by back-propagation algorithm to first
Each of model parameter is adjusted, and is then based on parameter building used car pricing model adjusted.
Method shown in the embodiment of the present application verifies the first pricing model of building, such as first pricing model
The checkout price of calculating and the concluded price of verification used car are inconsistent, pass through the conclusion of the business of the checkout price and verification used car
The difference of price adjusts first pricing model.Used car pricing model is ultimately generated according to adjustment result.
The used car pricing model that the embodiment of the present application shows method foundation can accurately predict the price of used car.
Embodiment 5:
In order to further increase the accuracy that the embodiment of the present application shows the used car pricing model of method building, the application
Implementation exemplifies a kind of method of adjustment of intermediate pricing model, specifically, please referring to Fig. 4:
Embodiment 5 and the technical solution shown in embodiment 4 there is similar step only difference is that, embodiment 4 is shown
Step S10315 in technical solution out the following steps are included:
S103151 calculates the checkout price and verifies the difference of the concluded price of used car, adjusts first price
The parameter of model obtains intermediate pricing model；
S103152 calculates the tune inspection price for verifying second-hand vehicle model according to the intermediate pricing model；
S103153 judges that described adjust examines price whether consistent with the concluded price of verification used car；
It is worth noting that, shown in the embodiment of the present application tune examine price and verify used car concluded price whether one
It causes, and not merely refers to the consistent of number, also refer to the infinite approach adjusted and examine price with the concluded price for verifying used car；Such as it adjusts
The difference examined price and verify the concluded price of used car is less than a certain specific threshold value, so that it may think the tune inspection valence
Lattice are consistent with the verification concluded price of used car.
If inconsistent, S103154 continues to adjust the parameter of intermediate pricing model, examines price and verification second-hand until adjusting
The concluded price of vehicle is consistent.
The embodiment of the present application obtains the prediction of used car from the input of vehicle characteristics by the full Connection Neural Network of definition
The output of price.Multiple neurons are united first, obtain the neuroid of multilayer, every layer of neuroid
Output is exactly the input of next layer of neuroid, until the feature of the output building of the last layer neuron, based on building
Feature generates intermediate pricing model.In order to guarantee that the accuracy of intermediate pricing model, the embodiment of the present application show method, pass through
Predicted value (adjust and examine price) compares the gap obtained between the two with true value (knock-down price of verification used car), then passes through
Back-propagation algorithm is adjusted each parameter of intermediate pricing model, generates new model, then proceedes to verify, constantly
The iteration above process examines price and the concluded price of verification used car to restrain until adjusting.
Repeatedly verification of the method by intermediate pricing model parameter shown in the embodiment of the present application, final guarantee to construct two
Handcart pricing model accurately predicts the price of used car.
Embodiment 6:
In order to improve the generating rate of used car pricing model, the embodiment of the present application shows a kind of method of adjustment of parameter,
Specifically, please referring to Fig. 5:
Technical solution shown in embodiment 6 has similar step to the technical solution shown in embodiment 5, unique to distinguish
Be step S103154 in the technical solution shown in embodiment 5 the following steps are included:
S1031541 determines the adjustment number of the intermediate pricing model；
S1031542 determines the adjustment amplitude of parameter according to the adjustment number；
When usually adjusting in the early stage, there are larger for the intermediate pricing model of the foundation trading environment actually located with used car
Discrepancy, adjust number numerical value it is smaller when, the parameter of intermediate pricing model is adjusted with biggish adjustment stride:
For example:
A is a parameter in intermediate pricing model, and A is 0.8 in the first pricing model, is adjusted for the first time, the adjustment of A
Amplitude is 0.5, and second of adjustment, the adjustment amplitude of A is 0.3, and third time adjusts, and the adjustment amplitude of A is 0.1；It adjusts for the first time,
The adjustment amplitude of A is 0.05；
What the embodiment of the present application was only exemplary shows the relationship of a kind of adjustment number and adjustment amplitude, in practical application
All adjustment amplitudes increase and reduced adjustment relationship with adjustment number, for example, specific dampings are waited, equal difference decaying, exponential damping,
In the protection scope of the technical solution shown in the embodiment of the present application, here, just different one illustrating since length is limited
It is bright.
S1031543 adjusts the parameter of intermediate pricing model according to the adjustment amplitude.
Method shown in the embodiment of the present application, adjustment initial stage set biggish adjustment amplitude, guarantee training initial stage fast approaching
More excellent solution, while can guarantee that the phase does not have too great fluctuation process to model after training again, to be more nearly local optimum.
Embodiment 7:
By exponential attenuation method, adjustment amplitude is determined according to the adjustment number, the adjustment amplitude is with the adjustment
Number increases and reduces.
The exponential attenuation method determines adjustment amplitude, can both accelerate the training speed for training initial stage, model can have been allowed to exist
The training more excellent solution of initial stage fast approaching, while can guarantee that the phase does not have too great fluctuation process to model after training again, thus more adjunction
Nearly local optimum.
Embodiment 8:
There is over-fitting in the used car pricing model of the method building shown in the embodiment of the present application in order to prevent, this Shen
It please implement to exemplify a kind of processing method of regularization, specifically, please referring to Fig. 6:
Step S103 in technical solution shown in embodiment 3 the following steps are included:
S10321 successively calculates the neuron, obtains modeling data according to the full Connection Neural Network；
Priori is added in the modeling data by S10322, is reduced the space of the modeling data, is obtained Target Modeling data；
S10323 constructs used car pricing model according to the Target Modeling data.
The modeling data is being added priori, is reducing solution space by the method shown in the embodiment of the present application, and reduction finds out mistake
A possibility that misunderstanding, and then guarantee that the phenomenon that over-fitting occurs in the used car pricing model of the embodiment of the present application building.
Embodiment 9:
The embodiment of the present application second aspect shows a kind of used car pricing device based on full Connection Neural Network, specifically
Referring to Fig. 7, described device includes:
Acquiring unit 21 constructs neuron according to the vehicle characteristics for obtaining vehicle characteristics；
Construction unit 22, for constructing full Connection Neural Network according to the neuron；
It is fixed to generate used car for successively calculating the neuron according to the full Connection Neural Network for generation unit 23
Valence model；
Computing unit 24, for calculating the price of target used car according to the used car pricing model.
Embodiment 10:
Construction unit 22 described in technical solution shown in embodiment 9 is also used to:
By the vehicle characteristics, neuron is constructed according to following formula:
Z=a1w1+……akwk+……aKwK+b；
Wherein, α=σ (wTA+b)=σ (w1a1+w2a2+…+wkak+b)；
a1,a2,...akFor vehicle characteristics, w1,w2,...wkFor the weight of vehicle characteristics, b is that bias term is second-hand for adjusting
Vehicle pricing model；σ is that activation primitive is used to vehicle characteristics transforming to a new feature space by the character representation in former space.
Embodiment 11:
Referring to Fig. 8, embodiment 10 shows generation unit 23 described in technical solution includes:
First generation unit 2311 generates for according to the full Connection Neural Network, successively calculating the neuron
One pricing model；
Calculation of price unit 2312 is examined, for calculating the inspection valence of verification used car according to first pricing model
Lattice；
First judging unit 2313, for judging whether the checkout price and the concluded price for verifying used car are consistent,
Determination unit 2314, if determining that first pricing model is used car pricing model for consistent
Adjustment unit 2315 generates used car pricing model if adjusting first pricing model for inconsistent.
Embodiment 12:
Please refer to Fig. 9: embodiment 11 shows adjustment unit 2315 described in technical solution and includes:
Intermediate pricing model generation unit 23151, for calculating the checkout price and verifying the concluded price of used car
Difference, adjust the parameter of first pricing model, obtain intermediate pricing model；
Calculation of price unit 23152 is examined, for calculating the tune for verifying second-hand vehicle model according to the intermediate pricing model
Examine price；
Second judgment unit 23153, for judge it is described adjust examine price and verify used car concluded price whether one
It causes；
Continue adjustment unit 23154, if continuing the parameter for adjusting intermediate pricing model for inconsistent, until adjusting inspection
It is consistent with the verification concluded price of used car to test price.
Embodiment 13:
Referring to Fig. 10, the continuation adjustment unit 23154 includes: in technical solution shown in embodiment 12
Number determination unit 231541 is adjusted, for determining the adjustment number of the intermediate pricing model；
Adjustment amplitude determination unit 231542, for determining the adjustment amplitude of parameter according to the adjustment number；
Parameter adjustment unit 231543, for adjusting the parameter of intermediate pricing model according to the adjustment amplitude.
Embodiment 14:
Embodiment 13 is shown in technical solution, and the adjustment amplitude determination unit 231543 is also used to,
By exponential attenuation method, adjustment amplitude is determined according to the adjustment number, the adjustment amplitude is with the adjustment
Number increases and reduces.
Embodiment 15:
In technical solution shown in embodiment 10, the construction unit 23 includes:
Layer-by-layer computing unit 2321, for successively calculating the neuron, being built according to the full Connection Neural Network
Modulus evidence；
Adding unit 2322 reduces the space of the modeling data, obtains mesh for priori to be added in the modeling data
Mark modeling data；
Used car pricing model construction unit 2323, for constructing used car price mould according to the Target Modeling data
Type.
Embodiment 16:
The embodiment of the present application fourth aspect shows a kind of server, please refers to Figure 12 and includes:
One or more processors 41；
Memory 42, for storing one or more programs；
When one or more of programs are executed by one or more of processors 41, so that one or more of places
Manage the method that device 41 realizes the embodiment of the present application crucial point.
The embodiment of the present application shows a kind of used car pricing method based on full Connection Neural Network, device and system, institute
The method of stating includes: acquisition vehicle characteristics, constructs neuron according to the vehicle characteristics；According to the neuron, full connection is constructed
Neural network；According to the full Connection Neural Network, the neuron is successively calculated, generates used car pricing model；According to institute
Used car pricing model is stated, the price of target used car is calculated.Vehicle characteristics are passed through mind by the method shown in the embodiment of the present application
Full Connection Neural Network is constructed through metanetwork, based on the calculated result of the neuron, used car pricing model is generated, is connecting entirely
It connects in the building process of neural network, the method shown in the embodiment of the present application, using the building of multi-layer network, by successively special
Sign transformation, the character representation by vehicle characteristics in former space transform to a new feature space, the feature after ultimately generating building,
Feature construction used car pricing model after being then based on building, the feature after the building that the embodiment of the present application generates, various dimensions
Consider the contribution that vehicle characteristics are fixed a price in used car, it is fixed to generate used car using the feature after the building shown in the embodiment of the present application
Valence model, the result error predicted in the price-setting process of used car are small.
Those skilled in the art after considering the specification and implementing the invention disclosed here, will readily occur to its of the application
Its embodiment.This application is intended to cover any variations, uses, or adaptations of the application, these modifications, purposes or
Person's adaptive change follows the general principle of the application and including the undocumented common knowledge in the art of the application
Or conventional techniques.The description and examples are only to be considered as illustrative, and the true scope and spirit of the application are by following
Claim is pointed out.
It should be understood that the application is not limited to the precise structure that has been described above and shown in the drawings, and
And various modifications and changes may be made without departing from the scope thereof.Scope of the present application is only limited by the accompanying claims.
It is worth noting that, in the specific implementation, the application also provides a kind of computer storage medium, wherein the computer
Storage medium can be stored with program, which may include the service providing method or use of user identity provided by the present application when executing
Step some or all of in each embodiment of family register method.The storage medium can be magnetic disk, CD, read-only storage note
Recall body (English: read-only memory, abbreviation: ROM) or random access memory (English: random access
Memory, referred to as: RAM) etc..
It is required that those skilled in the art can be understood that the technology in the embodiment of the present application can add by software
The mode of general hardware platform realize.Based on this understanding, the technical solution in the embodiment of the present application substantially or
Say that the part that contributes to existing technology can be embodied in the form of software products, which can deposit
Storage is in storage medium, such as ROM/RAM, magnetic disk, CD, including some instructions are used so that computer equipment (can be with
It is personal computer, server or the network equipment etc.) execute certain part institutes of each embodiment of the application or embodiment
The method stated.
Same and similar part may refer to each other between each embodiment in this specification.Especially for user identity
Service providing apparatus or user's registration device embodiment for, since it is substantially similar to the method embodiment, thus description
Comparison it is simple, related place is referring to the explanation in embodiment of the method.
Above-described the application embodiment does not constitute the restriction to the application protection scope.
Those skilled in the art after considering the specification and implementing the invention disclosed here, will readily occur to its of the application
Its embodiment.This application is intended to cover any variations, uses, or adaptations of the application, these modifications, purposes or
Person's adaptive change follows the general principle of the application and including the undocumented common knowledge in the art of the application
Or conventional techniques.The description and examples are only to be considered as illustrative, and the true scope and spirit of the application are by following
Claim is pointed out.
It should be understood that the application is not limited to the precise structure that has been described above and shown in the drawings, and
And various modifications and changes may be made without departing from the scope thereof.Scope of the present application is only limited by the accompanying claims this
Field technical staff after considering the specification and implementing the invention disclosed here, will readily occur to other embodiment party of the invention
Case.This application is intended to cover any variations, uses, or adaptations of the invention, these modifications, purposes or adaptability
Variation follows general principle of the invention and including the undocumented common knowledge or usual skill in the art of the present invention
Art means.The description and examples are only to be considered as illustrative, and true scope and spirit of the invention are by following claim
It points out.
It should be understood that the present invention is not limited to the precise structure already described above and shown in the accompanying drawings, and
And various modifications and changes may be made without departing from the scope thereof.The scope of the present invention is limited only by the attached claims.
Claims (15)
1. a kind of used car pricing method based on full Connection Neural Network, which is characterized in that the described method includes:
Vehicle characteristics are obtained, construct neuron according to the vehicle characteristics；
According to the neuron, full Connection Neural Network is constructed；
According to the full Connection Neural Network, the neuron is successively calculated, generates used car pricing model；
According to the used car pricing model, the price of target used car is calculated.
2. the method according to claim 1, wherein the acquisition vehicle characteristics, according to the vehicle characteristics structure
The step of building neuron specifically:
By the vehicle characteristics, neuron is constructed according to following formula:
Z=a1w1+……akwk+……aKwK+b；
Wherein, α=σ (wTA+b)=σ (w1a1+w2a2+…+wkak+b)；
a1,a2,...akFor vehicle characteristics, w1,w2,...wkFor the weight of vehicle characteristics, b is that bias term is fixed for adjusting used car
Valence model；σ is that activation primitive is used to vehicle characteristics transforming to a new feature space by the character representation in former space.
3. according to the method described in claim 2, it is characterized in that, the full Connection Neural Network of the basis, successively calculate described in
Neuron, generate used car pricing model the step of include:
According to the full Connection Neural Network, the neuron is successively calculated, generates the first pricing model；
According to first pricing model, the inspection price of verification used car is calculated；
Judge whether the checkout price and the concluded price for verifying used car are consistent,
If consistent, determine that first pricing model is used car pricing model
If inconsistent, first pricing model is adjusted, generates used car pricing model.
4. according to the method described in claim 3, it is characterized in that, the first pricing model of the adjustment, generates used car price
The step of model includes:
It calculates the checkout price and verifies the difference of the concluded price of used car, adjust the parameter of first pricing model,
Obtain intermediate pricing model；
According to the intermediate pricing model, the tune inspection price for verifying second-hand vehicle model is calculated；
Judge that described adjust examines price whether consistent with the concluded price of verification used car；
If inconsistent, continue the parameter for adjusting intermediate pricing model, until adjusting the knock-down price examined price and verify used car
Lattice are consistent.
5. according to the method described in claim 4, it is characterized in that, described continue the step of adjusting the parameter of intermediate pricing model
Specifically:
Determine the adjustment number of the intermediate pricing model；
The adjustment amplitude of parameter is determined according to the adjustment number；
According to the adjustment amplitude, the parameter of intermediate pricing model is adjusted.
6. according to the method described in claim 5, it is characterized in that, described determine that the adjustment amplitude of parameter has according to adjustment number
Body are as follows:
By exponential attenuation method, adjustment amplitude is determined according to the adjustment number, the adjustment amplitude is with the adjustment number
Increase and reduces.
7. according to the method described in claim 2, it is characterized in that, the full Connection Neural Network of the basis, successively calculate described in
Neuron, generate used car pricing model the step of include:
According to the full Connection Neural Network, the neuron is successively calculated, modeling data is obtained；
Priori is added in the modeling data, the space of the modeling data is reduced, obtains Target Modeling data；
According to the Target Modeling data, used car pricing model is constructed.
8. a kind of used car pricing device based on full Connection Neural Network, which is characterized in that described device includes:
Acquiring unit constructs neuron according to the vehicle characteristics for obtaining vehicle characteristics；
Construction unit, for constructing full Connection Neural Network according to the neuron；
Generation unit generates used car price mould for successively calculating the neuron according to the full Connection Neural Network
Type；
Computing unit, for calculating the price of target used car according to the used car pricing model.
9. device according to claim 8, which is characterized in that the construction unit is also used to:
By the vehicle characteristics, neuron is constructed according to following formula:
Z=a1w1+……akwk+……aKwK+b；
Wherein, α=σ (wTA+b)=σ (w1a1+w2a2+…+wkak+b)；
a1,a2,...akFor vehicle characteristics, w1,w2,...wkFor the weight of vehicle characteristics, b is that bias term is fixed for adjusting used car
Valence model；σ is that activation primitive is used to vehicle characteristics transforming to a new feature space by the character representation in former space.
10. device according to claim 9, which is characterized in that the generation unit includes:
First generation unit generates the first price mould for successively calculating the neuron according to the full Connection Neural Network
Type；
Calculation of price unit is examined, for calculating the inspection price of verification used car according to first pricing model；
First judging unit, for judging whether the checkout price and the concluded price for verifying used car are consistent,
Determination unit, if determining that first pricing model is used car pricing model for consistent
Adjustment unit generates used car pricing model if adjusting first pricing model for inconsistent.
11. device according to claim 10, which is characterized in that the adjustment unit includes:
Intermediate pricing model generation unit, the difference of the concluded price for calculating the checkout price and verification used car, is adjusted
The parameter of whole first pricing model, obtains intermediate pricing model；
Calculation of price unit is examined, for according to the intermediate pricing model, calculating the tune inspection price for verifying second-hand vehicle model；
Second judgment unit, for judging that described adjust examines price whether consistent with the concluded price of verification used car；
Continue adjustment unit, if continuing the parameter for adjusting intermediate pricing model for inconsistent, examines price and school until adjusting
The concluded price for testing used car is consistent.
12. device according to claim 11, which is characterized in that the continuation adjustment unit includes:
Number determination unit is adjusted, for determining the adjustment number of the intermediate pricing model；
Adjustment amplitude determination unit, for determining the adjustment amplitude of parameter according to the adjustment number；
Parameter adjustment unit, for adjusting the parameter of intermediate pricing model according to the adjustment amplitude.
13. device according to claim 12, which is characterized in that the adjustment amplitude determination unit is also used to,
By exponential attenuation method, adjustment amplitude is determined according to the adjustment number, the adjustment amplitude is with the adjustment number
Increase and reduces.
14. device according to claim 9, which is characterized in that the construction unit includes:
Layer-by-layer computing unit, for successively calculating the neuron, obtaining modeling data according to the full Connection Neural Network；
Adding unit reduces the space of the modeling data, obtains Target Modeling number for priori to be added in the modeling data
According to；
Used car pricing model construction unit, for constructing used car pricing model according to the Target Modeling data.
15. a kind of used car pricing system based on full Connection Neural Network, which is characterized in that the system comprises: application is flat
Platform server, the data storage server being connected with the application platform server, the data storage server setting exist
It inside the Platform Server or is independently arranged, the application platform server is connected by internet with terminal；
The terminal is used for the display of target used car price；
The storage that related data is used for according to storage server；
The application platform server is for realizing method such as of any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810201233.6A CN110264229A (en) | 2018-03-12 | 2018-03-12 | Used car pricing method based on full Connection Neural Network, device and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810201233.6A CN110264229A (en) | 2018-03-12 | 2018-03-12 | Used car pricing method based on full Connection Neural Network, device and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110264229A true CN110264229A (en) | 2019-09-20 |
Family
ID=67911536
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810201233.6A Pending CN110264229A (en) | 2018-03-12 | 2018-03-12 | Used car pricing method based on full Connection Neural Network, device and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110264229A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111598620A (en) * | 2020-05-20 | 2020-08-28 | 上海贤合汽车信息咨询服务有限公司 | Remote accurate pricing system for second-hand vehicles |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105787510A (en) * | 2016-02-26 | 2016-07-20 | 华东理工大学 | System and method for realizing subway scene classification based on deep learning |
| CN107145977A (en) * | 2017-04-28 | 2017-09-08 | 电子科技大学 | A kind of method that structured attributes deduction is carried out to online social network user |
| CN107315571A (en) * | 2016-04-27 | 2017-11-03 | 北京中科寒武纪科技有限公司 | A kind of apparatus and method for performing full articulamentum neutral net forward operation |
| CN107369043A (en) * | 2017-07-19 | 2017-11-21 | 河海大学常州校区 | A kind of used car price evaluation optimized algorithm based on BP neural network |
| CN107590537A (en) * | 2016-07-08 | 2018-01-16 | 香港理工大学 | For constructing the granulation Forecasting Methodology in probabilistic forecasting section |
-
2018
- 2018-03-12 CN CN201810201233.6A patent/CN110264229A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105787510A (en) * | 2016-02-26 | 2016-07-20 | 华东理工大学 | System and method for realizing subway scene classification based on deep learning |
| CN107315571A (en) * | 2016-04-27 | 2017-11-03 | 北京中科寒武纪科技有限公司 | A kind of apparatus and method for performing full articulamentum neutral net forward operation |
| CN107590537A (en) * | 2016-07-08 | 2018-01-16 | 香港理工大学 | For constructing the granulation Forecasting Methodology in probabilistic forecasting section |
| CN107145977A (en) * | 2017-04-28 | 2017-09-08 | 电子科技大学 | A kind of method that structured attributes deduction is carried out to online social network user |
| CN107369043A (en) * | 2017-07-19 | 2017-11-21 | 河海大学常州校区 | A kind of used car price evaluation optimized algorithm based on BP neural network |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111598620A (en) * | 2020-05-20 | 2020-08-28 | 上海贤合汽车信息咨询服务有限公司 | Remote accurate pricing system for second-hand vehicles |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Jiao et al. | Toward an automated auction framework for wireless federated learning services market | |
| CN113064700B (en) | Container deployment method and device | |
| CN109951333A (en) | Trust evaluation device based on subjective logic in the processing of edge calculations network video | |
| CN110383794A (en) | The system and method for application client are updated using multiple content delivery networks | |
| Reggiani et al. | Neural networks and logit models applied to commuters’ mobility in the metropolitan area of Milan | |
| Djedovic et al. | Innovative approach in modeling business processes with a focus on improving the allocation of human resources | |
| US11861664B2 (en) | Keyword bids determined from sparse data | |
| CN107194721A (en) | Service recommendation person based on reputation record analysis has found method | |
| CN110070452A (en) | Model training method, calculates equipment and computer readable storage medium at device | |
| Wang et al. | A reputation bootstrapping model for e-commerce based on fuzzy dematel method and neural network | |
| Chainakun et al. | Predicting the black hole mass and correlations in X-ray reverberating AGNs using neural networks | |
| Qie et al. | A stage model for agent-based emotional persuasion with an adaptive target: From a social exchange perspective | |
| CN110264229A (en) | Used car pricing method based on full Connection Neural Network, device and system | |
| CN110322342A (en) | Borrow or lend money construction method, system and the debt-credit Risk Forecast Method of risk forecast model | |
| CN113763928A (en) | Audio category prediction method and device, storage medium and electronic equipment | |
| CN117113222A (en) | Data analysis model generation method and device and electronic equipment | |
| Qu | Credible service selection in cloud environments | |
| Murray | The state of evaluation tools and systems for nonprofit organizations | |
| López-Martín et al. | Support vector regression for predicting the productivity of higher education graduate students from individually developed software projects | |
| Fan et al. | CRL-MABA: A Completion Rate Learning Based Accurate Data Collection Scheme in Large-Scale Energy Internet | |
| US20230012458A1 (en) | Identifying transaction processing retry attempts based on machine learning models for transaction success | |
| KR102360384B1 (en) | System for providing bigdata based reservation price probability distribution validation service for procurement auction | |
| CN114969209A (en) | Training method and device, and method and device for predicting resource consumption | |
| Sakhrawi et al. | Software enhancement effort estimation using machine learning regression methods | |
| Jiang et al. | User recruitment algorithm for maximizing quality under limited budget in mobile crowdsensing |
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 | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20200515 Address after: Room 323605, building 5, yard 1, Futong East Street, Chaoyang District, Beijing 100102 Applicant after: Youxuan (Beijing) Information Technology Co., Ltd Address before: Room 368, Room 302, No. 211 North Fute Road, China (Shanghai) Free Trade Pilot Area, Pudong New Area, Shanghai, 201315 Applicant before: Youestimate (Shanghai) Information Technology Co., Ltd |
|
| TA01 | Transfer of patent application right | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190920 |
|
| RJ01 | Rejection of invention patent application after publication |