CN115310256A - Vehicle estimation method and device - Google Patents
Vehicle estimation method and device Download PDFInfo
- Publication number
- CN115310256A CN115310256A CN202210719439.4A CN202210719439A CN115310256A CN 115310256 A CN115310256 A CN 115310256A CN 202210719439 A CN202210719439 A CN 202210719439A CN 115310256 A CN115310256 A CN 115310256A
- Authority
- CN
- China
- Prior art keywords
- attribute
- information
- vehicle
- estimation
- value
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/02—Reliability analysis or reliability optimisation; Failure analysis, e.g. worst case scenario performance, failure mode and effects analysis [FMEA]
Abstract
The invention relates to a vehicle valuation method and a device, wherein the method comprises the following steps: acquiring vehicle type information of a vehicle to be evaluated and attribute information related to evaluation of the vehicle to be evaluated; for each type of vehicle information, determining attribute categories corresponding to each attribute information and labeling the attribute categories to the attribute information to obtain labeled attribute information; inputting the marked attribute information and the vehicle type information into a target vehicle estimation model; and obtaining an estimation result of the vehicle to be estimated according to the labeling attribute information, the vehicle type information, the estimation function and the weight of the vehicle to be estimated. According to the scheme, from the perspective of the life cycle of the vehicle, the influence factors of the vehicle estimation are divided into the stable factors and the fluctuation factors, then estimation functions corresponding to all the factors are constructed, data of all dimensions of the vehicle can be standardized, vehicle type information and attribute information are introduced, the real value of the vehicle can be reflected, and the estimation accuracy is improved.
Description
Technical Field
The invention relates to the technical field of computers, in particular to a vehicle estimation method and device.
Background
With the change of domestic consumption level and consumption concept, the second-hand car market is gradually prosperous, and the transaction amount is steadily improved. It is increasingly important to provide accurate and reasonable estimates of used cars. It is more so today that evaluators are relied upon to give an estimate of used cars. However, due to the large differences in the experience and ability of evaluators, the same vehicle may receive different estimates via different evaluators. And because the individual evaluators cannot collect enough vehicle transaction data in the past as a reference to evaluate the current vehicle situation, the evaluators are not comprehensive in knowledge of the vehicle and the evaluation is not accurate.
Therefore, the existing second-hand vehicle estimation method has the problems that no unified estimation standard exists and estimation is not accurate enough.
Disclosure of Invention
The present invention is directed to solving at least one of the problems of the prior art. To this end, a first aspect of the invention proposes a vehicle estimation method, the method comprising:
acquiring attribute information of a vehicle to be evaluated;
determining attribute categories corresponding to the attribute information, wherein the attribute category columns are determined according to the influence degree of the attribute information on the vehicle estimation along with the time;
inputting the attribute information and the attribute category into a target vehicle estimation model; the estimation function in the estimation model of the target vehicle comprises sub estimation functions corresponding to the attribute information and weights corresponding to the attribute types; the target vehicle estimation model is constructed according to vehicle estimation data in a historical time period;
and obtaining an estimation result of the vehicle to be estimated according to the attribute information and the estimation function.
Optionally, the attribute category includes a stable attribute and a fluctuating attribute, the stable attribute is an attribute whose influence value on the vehicle estimation value over time is smaller than a first threshold, and the fluctuating attribute is an attribute whose influence value on the vehicle estimation value over time is greater than or equal to the first threshold.
Optionally, the target vehicle estimation model is obtained by training through the following method:
obtaining sample vehicle estimation data over a preset historical time period, the sample vehicle estimation data comprising: sample attribute information of a vehicle, an influence value of the sample attribute information of each sub-period on an estimation value of the vehicle; the preset historical time period comprises a plurality of sub-time periods;
respectively obtaining estimated value data corresponding to each stable attribute and estimated value data corresponding to each fluctuation attribute from the sample vehicle estimated value data to obtain a plurality of stable attribute estimated value data and a plurality of fluctuation attribute estimated value data;
respectively determining a first initial valuation function corresponding to each stability attribute according to an influence value in the stability attribute valuation data; respectively determining a second initial valuation function corresponding to each fluctuation attribute according to an influence value in the fluctuation attribute valuation data;
respectively training the first initial estimation function and the second initial estimation function by using the sample vehicle estimation data to obtain a stable attribute initial model formed by the first initial estimation function of each stable attribute and a fluctuation attribute initial model formed by the second initial estimation function of each fluctuation attribute;
and obtaining a target vehicle estimation model according to the preset weights of the stable attribute initial model and the fluctuation attribute initial model.
Optionally, the determining, according to the influence value in the stability attribute evaluation data, a first initial evaluation function corresponding to each stability attribute respectively includes:
setting the weight corresponding to each sub-time period;
for each stable attribute, calculating the influence value corresponding to each sub-time period and the weighted average value of the weights to obtain a weighted influence value;
and respectively determining the functional relationship between the vehicle age information and the weighted influence value under various value conditions of the stable attribute to obtain an initial evaluation function corresponding to the stable attribute.
Optionally, the sample attribute information includes optional configuration information, and before setting the weight corresponding to each sub-time period, the method further includes:
clustering the selected vehicle configuration information of each sample vehicle estimation data to obtain multiple groups of similar selected vehicle configuration information;
the determining a functional relationship between the age information and the weighted impact value comprises:
and respectively determining the functional relationship between the vehicle age information and the weighted influence value for each group of the similar selected configuration information.
Optionally, the determining, according to the influence value in the fluctuation attribute estimation data, a second initial estimation function corresponding to each fluctuation attribute respectively includes:
for each fluctuation attribute, taking the value of the fluctuation attribute as a feature vector, taking the influence value corresponding to the fluctuation attribute as a target value, and training a preset algorithm model to obtain a functional relation between the fluctuation attribute and the influence value;
and determining an initial estimation function corresponding to the fluctuation attribute according to the functional relation.
Optionally, the fluctuation attribute includes vehicle condition information, and the vehicle condition information includes: before training a preset algorithm model by using the value of the fluctuation attribute as a feature vector and the influence value corresponding to the fluctuation attribute as a target value, the accident information and the maintenance information of the vehicle further include:
obtaining vehicles with the same vehicle age information in the sample vehicle estimation data to obtain a data set of a plurality of vehicles with the same vehicle age information;
clustering the accident information of the vehicles in the data sets of the vehicles with the same vehicle age information to obtain a plurality of groups of similar accident information sets;
clustering the maintenance information of the vehicles in the accident information set to obtain a plurality of groups of similar maintenance information sets;
and acquiring the influence value of the vehicle in each group of the maintenance information sets.
Optionally, the training of the preset algorithm model with the value of the fluctuation attribute as a feature vector and the influence value corresponding to the fluctuation attribute as a target value includes:
for each maintenance information set, taking the age information, the accident information and the maintenance information as characteristic vectors, taking the influence value of the vehicle in the maintenance information set as a target value, and training a preset algorithm model to obtain a functional relation between the age information and the influence value corresponding to the maintenance information set;
the function relationship is a piecewise function which takes the vehicle age information as an independent variable and the influence value as a dependent variable, each subsection of the piecewise function is a constant function, and the value of the constant function is reduced along with the increase of the vehicle age information.
Optionally, the fluctuation attribute includes mileage information, and for each type of the mileage information value, a function relationship between the age information corresponding to the mileage information value and the influence value is: and a linear function with the vehicle age information as an independent variable and the influence value as a dependent variable, wherein the slope of the linear function is a negative number.
Optionally, the fluctuation attribute is month information of the vehicle transaction, and the functional relationship between the month information and the influence value is: and a broken line type function which takes the month as an independent variable and the influence value as a dependent variable.
Optionally, the attribute information at least includes vehicle type information, optional assembly location information, license plate information, usage information, and external environment information of the vehicle to be evaluated.
Optionally, the license plate information in the stable attribute includes license plate information and license plate attribute information; the use information in the stable attribute comprises use property information and user passing frequency information; the external environment information in the stable attribute includes country-friendly policy information.
Optionally, the usage information in the fluctuation attribute includes vehicle age information, vehicle condition information, mileage information, and claim information, and the external policy information in the fluctuation attribute includes month information of the vehicle transaction.
An embodiment of the present invention further provides a vehicle estimation device, where the device includes:
the attribute information acquisition module is used for acquiring attribute information of the vehicle to be evaluated;
the attribute type determining module is used for determining the attribute type corresponding to each attribute information, and the attribute type column determines the influence degree of the attribute information on the vehicle estimation along with the time;
a data input module for inputting the attribute information and the attribute category into a target vehicle valuation model; the estimation function in the target vehicle estimation model comprises sub estimation functions corresponding to the attribute information and weights corresponding to the attribute types; the target vehicle estimation model is constructed according to vehicle estimation data in a historical time period;
and the estimation module is used for obtaining an estimation result of the vehicle to be estimated according to the attribute information and the estimation function.
Optionally, the apparatus further includes a model training module, where the model training module is configured to:
obtaining sample vehicle estimation data over a preset historical time period, the sample vehicle estimation data comprising: sample attribute information of a vehicle, and an influence value of the sample attribute information of each sub-period on an estimation value of the vehicle; the preset historical time period comprises a plurality of the sub-time periods;
respectively obtaining estimated value data corresponding to each stable attribute and estimated value data corresponding to each fluctuation attribute from the sample vehicle estimated value data to obtain a plurality of stable attribute estimated value data and a plurality of fluctuation attribute estimated value data;
respectively determining a first initial evaluation function corresponding to each stability attribute according to an influence value in the stability attribute evaluation data; respectively determining a second initial evaluation function corresponding to each fluctuation attribute according to the influence value in the fluctuation attribute evaluation data;
respectively training the first initial estimation function and the second initial estimation function by using the sample vehicle estimation data to obtain a stable attribute initial model formed by the first initial estimation function of each stable attribute and a fluctuation attribute initial model formed by the second initial estimation function of each fluctuation attribute;
and obtaining a target vehicle estimation model according to the preset weights of the stable attribute initial model and the fluctuation attribute initial model.
Optionally, the model training module is specifically configured to:
setting the weight corresponding to each sub-time period;
for each stable attribute, calculating the influence value corresponding to each sub-time period and the weighted average value of the weights to obtain a weighted influence value;
and respectively determining the functional relationship between the vehicle age information and the weighted influence value under various value conditions of the stable attribute to obtain an initial evaluation function corresponding to the stable attribute.
Optionally, the model training module is specifically configured to:
clustering the optional equipment configuration information of each sample vehicle estimation data to obtain a plurality of groups of similar optional equipment configuration information;
the determining a functional relationship between the age information and the weighted impact value comprises:
and respectively determining the functional relationship between the vehicle age information and the weighted influence value for each group of the similar selected configuration information.
Optionally, the model training module is specifically configured to:
for each fluctuation attribute, taking the value of the fluctuation attribute as a feature vector, taking the influence value corresponding to the fluctuation attribute as a target value, and training a preset algorithm model to obtain a functional relation between the fluctuation attribute and the influence value;
and determining an initial valuation function corresponding to the fluctuation attribute according to the functional relation.
Optionally, the model training module is specifically configured to:
obtaining vehicles with the same vehicle age information in the sample vehicle estimation data to obtain a data set of a plurality of vehicles with the same vehicle age information;
clustering the accident information of the vehicles in the data sets of the vehicles with the same vehicle age information to obtain a plurality of groups of similar accident information sets;
clustering the maintenance information of the vehicles in the accident information set to obtain a plurality of groups of similar maintenance information sets;
and acquiring the influence value of the vehicle in each group of the maintenance information sets.
Optionally, the model training module is specifically configured to:
for each maintenance information set, taking the vehicle age information, the accident information and the maintenance information as characteristic vectors, taking the influence value of the vehicle in the maintenance information set as a target value, and training a preset algorithm model to obtain a functional relation between the vehicle age information and the influence value corresponding to the maintenance information set;
the function relationship is a piecewise function which takes the vehicle age information as an independent variable and the influence value as a dependent variable, each subsection of the piecewise function is a constant function, and the value of the constant function is reduced along with the increase of the vehicle age information.
A third aspect of the present invention proposes an electronic device comprising a processor and a memory having stored therein at least one instruction, at least one program, set of codes or set of instructions, which is loaded and executed by the processor to implement the vehicle estimation method according to the first aspect.
A fourth aspect of the present invention proposes a computer-readable storage medium having stored therein at least one instruction, at least one program, a set of codes, or a set of instructions, which is loaded and executed by a processor to implement the vehicle estimation method according to the first aspect.
According to the specific embodiment provided by the invention, the invention has the following technical effects:
the vehicle estimation method provided by the embodiment of the invention obtains the attribute information of the vehicle to be estimated; determining attribute categories corresponding to the attribute information, wherein the attribute category columns are determined according to the influence degree of the attribute information on vehicle estimation along with the time; inputting the attribute information and attribute categories into a target vehicle valuation model; the estimation function in the target vehicle estimation model comprises sub estimation functions corresponding to all attribute information and weights corresponding to attribute categories; constructing a target vehicle estimation model according to vehicle estimation data in a historical time period; and obtaining an estimation result of the vehicle to be estimated according to the attribute information and the estimation function. According to the scheme, the vehicle valuation model is constructed according to historical data, the influence factors of the vehicle valuation are divided into the stable factors and the fluctuation factors from the perspective of the life cycle of the vehicle, then valuation functions corresponding to all the factors are constructed, data of all dimensions of the vehicle can be standardized, vehicle type information and attribute information are introduced, the real value of the vehicle can be reflected, and the valuation accuracy is improved.
Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.
Drawings
In order to more clearly illustrate the technical solution of the present invention, the drawings used in the description of the embodiment or the prior art will be briefly described below. It is obvious that the drawings in the following description are only some embodiments of the invention, and that for a person skilled in the art it is also possible to derive other drawings from these drawings without inventive effort.
FIG. 1 is a flow chart illustrating steps of a method for vehicle estimation according to an embodiment of the present invention;
FIG. 2 is a flowchart illustrating steps in a method for training a target vehicle estimation model according to an embodiment of the present invention;
FIG. 3 is a schematic diagram illustrating a functional relationship between the car age and the weighted influence value corresponding to the card information provided by the embodiment of the present invention;
fig. 4 is a schematic diagram illustrating a functional relationship between a vehicle age and a weighted influence value corresponding to license plate attribute information provided in an embodiment of the present invention;
FIG. 5 is a schematic diagram illustrating a functional relationship between the age of the vehicle and the weighted impact value corresponding to the usage property information according to the embodiment of the present invention;
fig. 6 is a schematic diagram of a functional relationship between the vehicle age and the weighted influence value corresponding to the passing times information provided in the embodiment of the present invention;
FIG. 7 is a schematic diagram of an initial valuation function corresponding to the age of a vehicle according to an embodiment of the present invention;
FIG. 8 is a schematic diagram of a functional relationship between vehicle conditions and influence values provided by an embodiment of the present invention;
FIG. 9 is a graphical illustration of a functional relationship between mileage and impact values provided by an embodiment of the present invention;
FIG. 10 is a schematic diagram illustrating a functional relationship between a month and an influence value of a vehicle transaction according to an embodiment of the present invention;
fig. 11 is a block diagram showing a configuration of a vehicle estimation device according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, are within the scope of the present invention.
The present specification provides method steps as described in the examples or flowcharts, but more or fewer steps may be included based on routine or non-invasive labor. In practice, the system or server product may be implemented in a sequential or parallel manner (e.g., parallel processor or multi-threaded environment) according to the embodiments or methods shown in the figures.
FIG. 1 is a flow chart of steps of a method for vehicle estimation according to an embodiment of the present invention. The method comprises the following steps:
and step 101, acquiring attribute information of a vehicle to be evaluated.
The life cycle of an automobile generally includes a sales link, an order link, a production link, a delivery link, a use link, a change-of-hand sales link, an after-sales link, and a scrapping link.
In the selling link, the salesperson introduces the relevant functions, characteristics, matching and the like of the vehicle and preferential policies of countries and enterprises to the customers; in the order link, a customer can select a vehicle type, a color and other optional devices according to own preference, and a dealer can also consider whether a certain preferential policy is given according to the sales condition and the national policy, and generate an order; in the production link, manufacturers can produce vehicles according to orders of customers, and if the precision of the process and the quality of parts in the production link are problematic, the vehicles can be returned to factories and claimed after delivery, so that the valuation of the vehicles can be reduced; in the delivery link, partial damage of the vehicle caused by logistics or storage can occur, so that the vehicle is returned to a factory or maintained; in the using process, a customer can simply maintain or repair the vehicle due to the personal driving habits and the vehicle consumption caused by normal driving, and vehicle accidents caused by other accidents can cause different degrees of influence on the estimation value of the vehicle; in the roll-to-roll selling link, the roll-to-roll vehicle can be found under the personal condition of a client, when the roll-to-roll vehicle is found, the estimation value of the vehicle can be influenced, and when the occurrence frequency of the roll-to-roll vehicle is more, the influence on the estimation value is larger. In the after-sale link, the vehicle is generally maintained daily and the accident vehicle is maintained, and different after-sale contents have different influences on the vehicle estimation value. And in the scrapping link, the scrapped vehicles are evaluated to be the lowest, and the value of selling the parts separately is possibly higher than that of the whole vehicle.
In each link of the whole life cycle of the vehicle, different factors influence the estimation of the vehicle. The attribute information of the vehicle to be estimated can be obtained from each link of the life cycle of the vehicle, and the attribute information is a factor influencing the estimation of the vehicle.
In one possible embodiment, the attribute information includes at least vehicle type information, optional equipment location information, license plate information, usage information, and external environment information of the vehicle to be evaluated.
Specifically, the vehicle type information refers to the license plate and model of the vehicle, and is a main factor influencing vehicle estimation. The optional configuration information refers to configuration which can be directly selected according to the preference or the requirement of a purchaser when ordering a vehicle. For example, navigation, leather seats, power skylights, vehicle body stabilizing systems and the like are added, and the comfort and the safety can be improved by adding optional configurations.
The license plate information indicates whether the vehicle has been registered.
The usage information is information related to the usage of the vehicle, such as whether the usage of the vehicle is personal or business, the number of times of passing a family of the vehicle, the mileage traveled by the vehicle, basic condition information of the vehicle, claim information of the vehicle, and the like.
The external environment information refers to external information that is not related to the vehicle itself but affects the vehicle estimation value, such as the month of selling the vehicle at present, a national preferential policy when selling the vehicle, and the like.
And 102, determining attribute types corresponding to the attribute information, wherein the attribute type column is determined according to the influence degree of the attribute information on the vehicle estimation value along with the time.
In each link of the whole life cycle of the vehicle, different factors influence the estimation of the vehicle, and the influence degree of different attribute information on the estimation changes along with the time. And classifying the attribute information according to the influence degree of each attribute information on the vehicle estimation value along with the time. Attribute information with a large influence degree on vehicle estimation values along with the time continuation is classified into one type, attribute information with a small influence degree is classified into another type, and attribute types corresponding to the attribute information are obtained.
For example, attribute information such as listing information, license plate attribute information, use property information, and the like has little influence on evaluation over time, and such attributes are classified into one category; the attribute information such as the vehicle age, the vehicle condition, the mileage, the claim and the like has a large influence on the evaluation value along with the time, and the attributes are classified into another class.
And combining order data of the vehicle, attribute information of the vehicle, estimation data of the vehicle, after-sales data and the like, constructing an initial estimation model according to the dimension of the attribute class, and training the initial estimation model according to the influence of each attribute class on estimation to obtain the estimation model of the target vehicle.
The target vehicle estimation model comprises estimation functions corresponding to all vehicle types, and the estimation functions are weighted summation of sub estimation functions corresponding to all attribute information and weights corresponding to attribute categories to which the sub estimation functions belong.
It should be noted that, the estimation of the vehicle is greatly influenced by the brand and model of the vehicle, and a corresponding target vehicle estimation model may be constructed for each vehicle type, and then when the vehicle estimation is performed, the attribute information and attribute category of the vehicle are input into the target vehicle estimation model of the corresponding vehicle type.
In one possible embodiment, the attribute category includes a stable attribute that is an attribute whose influence value on the vehicle estimation value over time is smaller than a first threshold value, and a fluctuating attribute that is an attribute whose influence value on the vehicle estimation value over time is greater than or equal to the first threshold value.
Specifically, the attribute information is divided according to the attribute categories of the stable attribute and the fluctuation attribute, and the attribute categories are labeled to the attribute information, so that the subsequent evaluation functions are conveniently and respectively constructed for the stable attribute and the fluctuation attribute.
Further, the target vehicle estimation model can be expressed by equation (1) as follows:
wherein, w i Weight representing the ith stability attribute, S i Represents the corresponding estimate of the ith volatility attribute, w j Weight, R, representing the jth fluctuation attribute j The evaluation value corresponding to the jth fluctuation attribute is shown, N is the number of stable attributes, and M is the number of fluctuation attributes.
In one possible embodiment, the license plate information in the stable attribute includes license plate information and license plate attribute information; the use information in the stable attribute comprises use property information and number of times of passing a house; the external environment information in the stable attribute includes country-friendly policy information.
The stable attribute is an attribute having a small influence on the estimation of the vehicle over time. Attribute information such as license plate information, license plate attribute information, usage property information, passing times information, and national preferential policy information has little influence on vehicle estimation according to changes in vehicle age, and therefore the attribute type of the attribute information is a stable attribute.
In one possible embodiment, the usage information in the fluctuation attributes includes vehicle age, vehicle condition, mileage, claims, and the external policy information in the fluctuation attributes includes the month of the vehicle transaction.
The fluctuation attribute is an attribute having a large influence on the evaluation with the lapse of time. The attribute information includes the age of the vehicle, the condition of the vehicle, the mileage, the claim, and the month of the vehicle transaction, and the attribute category of the attribute information is a fluctuation attribute because the attribute information has a large influence on the vehicle estimation value with the time.
And step 104, obtaining an estimation result of the vehicle to be estimated according to the attribute information and the estimation function.
Each vehicle type corresponds to a target evaluation function, and each target evaluation function comprises a plurality of sub evaluation functions corresponding to a plurality of attribute information. And substituting the labeled attribute information into the corresponding sub-estimation function to obtain a sub-estimation result. For example, the attribute information includes license plate attribute information, and the license plate attribute information is substituted into a first sub-evaluation function corresponding to the license plate attribute information to obtain a first sub-evaluation result for the license plate attribute information. And the attribute information also comprises mileage information, and the mileage information is substituted into a second sub-evaluation function corresponding to the mileage information to obtain a second sub-evaluation result aiming at the mileage information.
Thus, a plurality of sub-evaluation results corresponding to the plurality of labeling attribute information are obtained.
And carrying out weighted summation on the plurality of sub-estimation results and the weights corresponding to the attribute categories. For example, if the weight of the stable attribute is 0.4 and the weight of the fluctuating attribute is 0.6, the first sub-estimation result is multiplied by the weight of the stable attribute of 0.4, the second sub-estimation result is multiplied by the weight of the fluctuating attribute of 0.6, and the multiplication results are summed up to obtain the estimation result of the vehicle.
The weights for the stable attribute and the fluctuating attribute may be preset empirically. Different weights are set for the stable attribute and the fluctuation attribute, and the influence degree of each attribute information on the estimation result can be scientifically evaluated, so that the estimation result is more accurate.
In summary, in the embodiment of the present invention, attribute information of a vehicle to be evaluated is obtained; determining attribute categories corresponding to the attribute information, wherein the attribute category columns are determined according to the influence degree of the attribute information on the vehicle estimation along with the time; inputting the attribute information and the attribute category into a target vehicle estimation model; the estimation function in the target vehicle estimation model comprises sub estimation functions corresponding to the attribute information and weights corresponding to the attribute types; the target vehicle estimation model is constructed according to vehicle estimation data in a historical time period; and obtaining an estimation result of the vehicle to be estimated according to the attribute information and the estimation function. According to the scheme, from the perspective of the life cycle of the vehicle, the influence degree of the attribute information of the vehicle on the vehicle estimation is classified according to the time duration, and then the estimation function corresponding to each attribute information and the weight corresponding to the attribute category are constructed, so that the attribute information of each dimension of the vehicle is integrated, the estimation algorithm is standardized, the real value of the vehicle is reflected, and the estimation accuracy is improved.
FIG. 2 is a flowchart illustrating steps of a method for training a target vehicle estimation model according to an embodiment of the present invention. The method comprises the following steps:
The preset historical time period may be a time period before the current time, for example, ten years, three months, etc. before the current time, and may be specifically preset according to the needs and data conditions.
The sub-period refers to a period within a preset historical period, for example, if the preset historical period is ten years before the current time, the sub-period may be the first year, the second year, the third year, etc. before the current time.
The vehicle estimation data within the preset historical time period can be obtained from the database to obtain sample vehicle estimation data. The sample vehicle estimation value data may specifically include sample attribute information of the vehicle, an influence value of the sample attribute information of each sub-period on the vehicle estimation value, and the like.
The influence value refers to an influence value of the sample attribute information on the vehicle estimation value. For example, for a first sample vehicle, if the card information is a card, the impact value is the impact value of the card on the vehicle estimate, which may be 0.9, indicating that the card reduces the vehicle estimate to 0.9 times the original price; for the second sample vehicle, if the hit card information is a miss, the impact value is the impact of the miss on the vehicle's estimate, which may be 1.1, indicating that the miss increases the vehicle's estimate by a factor of 1.1 times the original price.
The stability attribute estimation data includes stability attribute information and corresponding influence values, and the fluctuation attribute estimation data includes fluctuation attribute information and corresponding influence values.
And taking the stable attribute information as a function variable, taking the influence value as a function value, selecting a proper algorithm according to the data characteristics of the stable attribute, and constructing an evaluation function corresponding to the stable attribute to obtain a first initial evaluation function.
And (4) similarly, taking the fluctuation attribute information as a function variable, taking the influence value as a function value, selecting a proper algorithm according to the data characteristics of the fluctuation attribute, and constructing an evaluation function corresponding to the fluctuation attribute to obtain a second initial evaluation function.
Because the data characteristics of each attribute information are different, different algorithm models can be selected to construct the initial valuation function. For example, for the option configuration information, which includes various categories of configuration information, a classification algorithm, such as an SVM (support Vector machine MacS) algorithm or a decision tree algorithm, may be used, the feature Vector is the option configuration data, and the target result is the estimation, so as to construct the initial estimation function.
It should be noted that the stable attribute or the fluctuating attribute includes a plurality of attribute information, and each attribute information corresponds to an initial valuation function.
First, a first algorithm model and a second algorithm model are obtained by using a first initial estimation function and a second initial estimation function. And for each stable attribute, inputting the stable attribute estimation data into the first algorithm model to obtain a predicted value, comparing the predicted value with an influence value corresponding to the stable attribute to obtain a residual value, modifying parameters of the first algorithm model according to the residual value, and continuing training to obtain a first initial model corresponding to the stable attribute.
And combining the first initial models corresponding to the stable attributes to obtain the stable attribute initial models.
And after the training is finished, combining the second initial models corresponding to the fluctuation attributes to obtain the fluctuation attribute initial model.
And step 205, obtaining an estimation model of the target vehicle according to preset weights of the stable attribute initial model and the fluctuation attribute initial model.
Further, a first weight of a stable attribute and a second weight of a fluctuating attribute which are preset are distributed to the two initial models to obtain a target vehicle estimation model.
In a possible implementation manner, the determining the first initial estimation function corresponding to each of the stable attributes according to the influence value in the stable attribute estimation data includes steps 2031 to 2033:
step 2031, setting weights corresponding to the sub-time periods;
step 2032, for each of the stable attributes, calculating a weighted average of the weight and the influence value corresponding to each of the sub-periods to obtain a weighted influence value;
step 2033, determining a functional relationship between the vehicle age information and the weighted influence value under various value-taking conditions of the stable attribute, respectively, to obtain an initial valuation function corresponding to the stable attribute.
In steps 2031 to 2033, since the referential property of the data of the closer historical time is relatively higher and the referential property of the data of the farther historical time is relatively lower, a higher weight may be set for the closer sub-period and a lower weight may be set for the farther sub-period. For example, if the preset historical period is ten years before the current time and the sub-period is each of ten years, the weights of 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0 are assigned in order from far to near for each of ten years.
And acquiring the influence value corresponding to each sub-time period for each type of sample attribute information. For example, the impact value of each year of a decade. And then carrying out weighted average on the influence value of each year and the current weight, and constructing a functional relation between the vehicle age and the weighted influence value to further obtain an initial estimation function corresponding to the sample attribute information.
For each stability factor, the estimation is influenced by two factors, namely the age of the vehicle and the stability attribute, and the functional relationship is obtained by determining the next value before obtaining the functional relationship between the other two factors. Therefore, the value of the stable attribute is fixed, and the functional relationship between the age of the vehicle and the influence value is determined. Therefore, the functional relationship between the age of the vehicle and the weighted influence value under various value taking conditions of the stability attribute is respectively determined, and an initial valuation function corresponding to the stability attribute is obtained.
For example, if the sample attribute information is the card information, the method for constructing the initial valuation function corresponding to the sample attribute information is as follows:
calculating the influence value and the weighted influence value of the weight corresponding to the cards in each sub-time period, as shown in table 1:
TABLE 1 influence values and weights of tiles
| Sub-period of time | Influence value of putting cards | Weight of | Mean value weight |
| 2021 year old | 0.875 | 0.1 | 0.0875 |
| 2020 to | 0.825 | 0.2 | 0.165 |
| 2019 | 0.7875 | 0.3 | 0.23625 |
| 2018 years old | 0.7875 | 0.4 | 0.315 |
| 2017 | 0.8625 | 0.5 | 0.43125 |
| 2016 (year) | 0.75 | 0.6 | 0.45 |
| 2015 years | 0.7875 | 0.7 | 0.55125 |
| 2014 | 0.825 | 0.8 | 0.66 |
| 2013 | 0.7875 | 0.9 | 0.70875 |
| 2012 year | 0.8375 | 1.0 | 0.8375 |
In table 1, the weighted impact values: sum (mean weight)/sum (weight) =0.807727.
Similarly, the influence value corresponding to the non-hit of each sub-period is determined, and the weighted influence value of the influence value and the weight corresponding to the non-hit of each sub-period is calculated using the weights in table 1, for example, the result is 1.030.
Because the card-registering information is a stable attribute, the influence value of the card-registering information on the vehicle estimation value does not change along with the vehicle age, and therefore, the function relationship between the vehicle age and the weighted influence value is two constant functions.
Fig. 3 is a schematic diagram of a functional relationship between the car age and the weighted influence value corresponding to the card information provided by the embodiment of the present invention.
As shown in fig. 3, the function for the hit cards is y =0.807x, and the function for the miss cards is y =1.030x. Wherein x represents the age of the vehicle and y represents the influence value. The functional relation is the initial valuation function.
Fig. 4 is a schematic diagram illustrating a functional relationship between the vehicle age and the weighted influence value corresponding to the license plate attribute information according to the embodiment of the present invention.
As shown in fig. 4, when the license plate attribute is a business, the corresponding function is y =0.970x, and when the license plate attribute is an individual, the corresponding function is y =1.000x.
Fig. 5 is a schematic diagram of a functional relationship between the age of the vehicle and the weighted influence value corresponding to the usage property information according to the embodiment of the present invention.
As shown in fig. 5, the corresponding function is y =0.71x when the usage property is operation, and the corresponding function is y =1.00x when the usage property is non-business.
Fig. 6 is a schematic diagram of a functional relationship between the vehicle age and the weighted influence value corresponding to the number of passing households information provided by the embodiment of the invention.
As shown in fig. 6, when the number of house-crossings is one, the corresponding function is y = -0.4x, when the number of house-crossings is two, the corresponding function is y = -0.8x, when the number of house-crossings is three, the corresponding function is y = -1.2x, and when the number of house-crossings is four, the corresponding function is y = -1.6x.
In a possible implementation manner, the sample attribute information includes option configuration information, and before setting the weight corresponding to each of the sub-time periods, the method further includes:
clustering the selected vehicle configuration information of each sample vehicle estimation data to obtain multiple groups of similar selected vehicle configuration information;
the determining a functional relationship between the vehicle age information and the weighted impact value comprises:
and respectively determining the functional relationship between the vehicle age information and the weighted influence value for each group of the similar selected configuration information.
In the embodiment of the invention, the assembly information selection comprises various types such as increasing navigation, increasing leather seats, increasing power skylights, increasing a vehicle body stabilizing system and the like, and the various types of assembly information selection can be classified by adopting a clustering method to obtain multiple groups of similar assembly configuration information.
Thus, when the functional relationship is constructed, the corresponding functional relationship is constructed for each group of similar selected configuration information.
In a possible implementation manner, the determining the second initial estimation function corresponding to each fluctuation attribute according to the influence value in the fluctuation attribute estimation data includes the following steps 2034 to 2035:
step 2034, for each fluctuation attribute, taking the value of the fluctuation attribute as a feature vector, taking the influence value corresponding to the fluctuation attribute as a target value, and training a preset algorithm model to obtain a functional relationship between the fluctuation attribute and the influence value;
step 2035, determining an initial estimation function corresponding to the fluctuation attribute according to the functional relationship.
In steps 2034 to 2035, the influence value is the ratio between the average value of the used car transaction prices and the average value of the new car transaction prices for each vehicle type. For example, for the 001 model, the average value of the second hand car transaction price is 9 ten thousand, and the average value of the new car transaction price is 10 ten thousand, and the influence value is 9/10.
The fluctuation attributes include vehicle age, vehicle condition, mileage, claim, month of vehicle transaction, etc., and table 2 is a vehicle age data set provided by the embodiment of the present invention, taking vehicle age as an example.
TABLE 2 age of vehicle data set
| Vehicle model | Age of vehicle | Influence value |
| 001 | 2 | 0.92 |
| 002 | 4 | 0.88 |
| 003 | 1 | 0.91 |
| 004 | 5 | 0.86 |
| …… | …… | …… |
For each vehicle type, the vehicle age data set in table 2 is input into an algorithm model, and an initial estimation function corresponding to the vehicle age can be obtained through training.
Fig. 7 is a schematic diagram of an initial estimation function corresponding to the vehicle age according to an embodiment of the present invention.
Referring to fig. 7, as the age of the vehicle increases, the influence of the age of the vehicle on the vehicle estimation gradually decreases.
In one possible embodiment, the fluctuation attribute includes vehicle condition information including: before training a preset algorithm model by using the value of the fluctuation attribute as a characteristic vector and the influence value corresponding to the fluctuation attribute as a target value, the accident information and the maintenance information of the vehicle further comprise the following steps 301 to 304:
301, obtaining vehicles with the same vehicle age information in the sample vehicle estimation data to obtain a data set of a plurality of vehicles with the same vehicle age information;
step 302, clustering the accident information of the vehicles in the data sets of the vehicles with the same vehicle age information to obtain a plurality of groups of similar accident information sets;
303, clustering the maintenance information of the vehicles in the accident information set to obtain a plurality of groups of similar maintenance information sets;
and 304, acquiring the influence value of the vehicle in each group of the maintenance information set.
In steps 301-304, the vehicle condition has a step-like influence on the vehicle estimation, and during the life cycle of the vehicle, the vehicle is repaired by accidents at different time points, and the influence weight is gradually reduced along with the increase of time. The method comprises the steps of obtaining transaction data and maintenance record data in a preset historical time period, and extracting data of vehicle types, vehicle ages, parts, accident information, maintenance information and influence values. The parts comprise hundreds of engine room covers, front bumpers, front windshields, front fenders and the like and automobile core parts; accident types comprise dozens of types such as damage, claim, fault, corrosion, liquid leakage, oxidation and the like; the maintenance types include dozens of types such as paint spraying, replacing, sheet metal cutting and the like. The accident information and the repair information for each vehicle are one or more of the accident types and the repair types described above. And aligning each vehicle type according to the vehicle age, selecting vehicles with the same vehicle age, accident information and maintenance information, and calculating the influence of the vehicles on the estimation.
In a possible implementation manner, the training a preset algorithm model by using the value of the fluctuation attribute as a feature vector and the influence value corresponding to the fluctuation attribute as a target value includes:
for each maintenance information set, taking the vehicle age information, the accident information and the maintenance information as characteristic vectors, taking the influence value of the vehicle in the maintenance information set as a target value, and training a preset algorithm model to obtain a functional relation between the vehicle age information and the influence value corresponding to the maintenance information set; the function relationship is a piecewise function which takes the vehicle age information as an independent variable and the influence value as a dependent variable, each subsection of the piecewise function is a constant function, and the value of the constant function is reduced along with the increase of the vehicle age information.
In the embodiment of the invention, for vehicles with the same vehicle age, accident information and maintenance information, the vehicle age, the accident information and the maintenance information are used as characteristic vectors, the influence value of the vehicle in the maintenance information set is used as a target value, and a preset algorithm model is trained to obtain the functional relation between the vehicle age information and the influence value corresponding to each maintenance information set.
FIG. 8 is a schematic diagram of a functional relationship between vehicle conditions and impact values provided by an embodiment of the present invention.
As shown in fig. 8, since each maintenance information set includes different accident information and maintenance information, the function curves of the age and the influence value are different for the different maintenance information sets.
Similar to the stable attribute, for each fluctuating attribute, the estimation is influenced by two factors, namely the vehicle age and the fluctuating attribute, and the functional relationship needs to be obtained by determining the next factor before the functional relationship between the other two factors is obtained. Therefore, the value of the fluctuation attribute is fixed, and the functional relationship between the age of the vehicle and the estimation value (the estimation value is the influence value in the graph) is determined. Therefore, in fig. 8, the independent variable is the age of the vehicle, and the dependent variable is the influence value.
As can be seen from fig. 8, the function relationship is a piecewise function with the vehicle age information as an independent variable and the influence value as a dependent variable, each subsection of the piecewise function is a constant function, and the value of the constant function decreases as the vehicle age information increases.
In a possible implementation manner, the fluctuation attribute includes mileage information, and for each mileage information value, a functional relationship between the age information and the influence value corresponding to the mileage information value is: and a linear function with the vehicle age information as an independent variable and the influence value as a dependent variable, wherein the slope of the linear function is a negative number.
Specifically, the estimation value of the vehicle gradually decreases as the mileage increases. Table 3 is a mileage dataset provided by an embodiment of the present invention.
TABLE 3 Mileage dataset
| Vehicle model | Mileage (Km) | Influence value |
| 001 | 800 | 0.93 |
| 002 | 4000 | 0.89 |
| 003 | 2000 | 0.91 |
| 004 | 6000 | 0.87 |
| …… | …… | …… |
Fig. 9 is a schematic diagram of a functional relationship between mileage and an influence value according to an embodiment of the present invention.
Similar to fig. 8, the value of the volatility attribute (i.e., mileage) is first fixed and a functional relationship between the age of the vehicle and the impact value is determined. As can be seen from fig. 9, the function relationship between the vehicle age and the influence value corresponding to the mileage value is a linear function, and the slope of the linear function is a negative number, that is, the larger the mileage value is, the smaller the influence of the vehicle age on the vehicle estimation value is.
In one possible embodiment, the fluctuation attribute is month information of the vehicle transaction, and the functional relationship between the month information and the influence value is: and a broken line type function which takes the month as an independent variable and the influence value as a dependent variable.
Since sales may vary in short and busy seasons as months change, and thus may result in sales at different time points, the corresponding valuations may also vary. Acquiring transaction data of nearly five years, extracting vehicle type, closing month and residual value influence characteristic data, dividing one year into twelve months, and aligning the vehicle closing data according to the month of closing time. For one vehicle model, a weighted average of its impact on vehicle residuals in the same month is calculated. The calculation method is that the data of the same month in the last five years are weighted and averaged, and the weight is distributed according to [0.1,1.0], so that the influence of the vehicle type on the estimation value in each month is obtained.
FIG. 10 is a schematic diagram of the functional relationship between the month of the vehicle transaction and the impact value provided by the embodiment of the invention.
It can be seen that the function relationship is a broken line type function with month as an argument and an influence value as a dependent variable.
In addition, claims may arise from product quality issues for automobile manufacturers throughout the life of the vehicle. And acquiring second-hand vehicle transaction data and claim data in a preset historical time period, and extracting vehicle types, claim items and influence values. As claims increase with age, the impact on valuation decreases. And for each vehicle type, training to obtain a claim regression model according to the claim term characteristics and the influence values by using a prediction class algorithm such as Lasso regression.
In summary, the target vehicle estimation model training method provided by the embodiment of the invention introduces various estimation influence factors such as vehicle age, mileage, month, preference, vehicle conditions and the like, is beneficial to calculating the estimation of the vehicle from multiple dimensions, can reflect the real estimation of the vehicle better due to the introduction of more sample vehicle estimation data, and improves the estimation accuracy.
Fig. 11 is a block diagram showing a configuration of a vehicle estimation device according to an embodiment of the present invention.
As shown in fig. 11, the vehicle estimation device 400 includes:
a vehicle type attribute information obtaining module 401, configured to obtain vehicle type information of a vehicle to be evaluated and attribute information related to evaluation of the vehicle to be evaluated;
a labeling attribute information determining module 402, configured to determine an attribute category corresponding to each attribute information and label the attribute category to the attribute information to obtain labeling attribute information; the attribute category comprises a stable attribute and a fluctuating attribute, the stable attribute is an attribute with invariable estimation value influence on the vehicle along with the time, and the fluctuating attribute is an attribute with variable influence on the estimation value along with the time;
an input module 403, configured to input the labeled attribute information and the vehicle type information into a target vehicle estimation model; the target vehicle estimation model comprises estimation functions corresponding to the vehicle type information, and the estimation functions comprise sub estimation functions corresponding to the attribute information and weights corresponding to the attribute categories; the target vehicle estimation model is constructed according to vehicle estimation data in a historical time period;
an estimation result determining module 404, configured to obtain an estimation result of the vehicle to be estimated according to the labeling attribute information, the vehicle type information, the estimation function, and the weight of the vehicle to be estimated.
It can be clearly understood by those skilled in the art that, for convenience and simplicity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
In yet another embodiment provided by the present invention, there is also provided an apparatus comprising a processor and a memory having stored therein at least one instruction, at least one program, set of codes, or set of instructions that is loaded and executed by the processor to implement the vehicle estimation method described in an embodiment of the present invention.
In yet another embodiment provided by the present invention, a computer-readable storage medium is also provided, having stored therein at least one instruction, at least one program, set of codes, or set of instructions, which is loaded and executed by a processor to implement the vehicle estimation method described in the embodiments of the present invention.
In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed on a computer, cause the processes or functions described in accordance with the embodiments of the invention to occur, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that incorporates one or more of the available media. The usable medium may be a magnetic medium (e.g., floppy Disk, hard Disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid State Disk (SSD)), among others.
It should be noted that, in this document, relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrases "comprising one of 8230; \8230;" 8230; "does not exclude the presence of additional like elements in a process, method, article, or apparatus that comprises the element.
All the embodiments in the present specification are described in a related manner, and the same and similar parts among the embodiments may be referred to each other, and each embodiment focuses on differences from other embodiments. In particular, for the system embodiment, since it is substantially similar to the method embodiment, the description is simple, and for the relevant points, reference may be made to the partial description of the method embodiment.
The above description is only for the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention shall fall within the protection scope of the present invention.
Claims (16)
1. A vehicle estimation method, characterized in that the method comprises:
acquiring attribute information of a vehicle to be evaluated;
determining attribute categories corresponding to the attribute information, wherein the attribute category columns are determined according to the influence degree of the attribute information on the vehicle estimation along with the time;
inputting the attribute information and the attribute category into a target vehicle estimation model; the estimation function in the target vehicle estimation model comprises sub estimation functions corresponding to the attribute information and weights corresponding to the attribute types; the target vehicle estimation model is constructed according to vehicle estimation data in a historical time period;
and obtaining an estimation result of the vehicle to be estimated according to the attribute information and the estimation function.
2. The method according to claim 1, wherein the attribute categories include a stable attribute that is an attribute whose influence value on the vehicle estimation value over time is smaller than a first threshold value, and a fluctuating attribute that is an attribute whose influence value on the vehicle estimation value over time is greater than or equal to the first threshold value.
3. The method of claim 2, wherein the target vehicle estimation model is trained by:
obtaining sample vehicle estimation data over a preset historical time period, the sample vehicle estimation data comprising: sample attribute information of a vehicle, an influence value of the sample attribute information of each sub-period on an estimation value of the vehicle; the preset historical time period comprises a plurality of sub-time periods;
respectively obtaining estimated value data corresponding to each stable attribute and estimated value data corresponding to each fluctuation attribute from the sample vehicle estimated value data to obtain a plurality of stable attribute estimated value data and a plurality of fluctuation attribute estimated value data;
respectively determining a first initial valuation function corresponding to each stability attribute according to an influence value in the stability attribute valuation data; respectively determining a second initial valuation function corresponding to each fluctuation attribute according to an influence value in the fluctuation attribute valuation data;
respectively training the first initial valuation function and the second initial valuation function by using the sample vehicle valuation data to obtain a stable attribute initial model formed by the first initial valuation function of each stable attribute and a fluctuation attribute initial model formed by the second initial valuation function of each fluctuation attribute;
and obtaining a target vehicle estimation model according to the preset weights of the stable attribute initial model and the fluctuation attribute initial model.
4. The method of claim 3, wherein said determining a first initial valuation function for each of said stable attributes based on impact values in said stable attribute valuation data comprises:
setting the weight corresponding to each sub-time period;
for each stable attribute, calculating a weighted average value of the influence value corresponding to each sub-time period and the weight to obtain a weighted influence value;
and respectively determining the functional relationship between the vehicle age information and the weighted influence value under various value-taking conditions of the stable attribute to obtain an initial valuation function corresponding to the stable attribute.
5. The method according to claim 4, wherein the sample attribute information includes option configuration information, and before setting the weight corresponding to each of the sub-periods, the method further includes:
clustering the selected vehicle configuration information of each sample vehicle estimation data to obtain multiple groups of similar selected vehicle configuration information;
the determining the functional relationship between the vehicle age information and the weighted influence value comprises:
and respectively determining the functional relationship between the vehicle age information and the weighted influence value for each group of the similar selected configuration information.
6. The method of claim 3, wherein said determining a second initial valuation function for each of said volatility attributes based on impact values in said volatility attribute valuation data comprises:
for each fluctuation attribute, taking the value of the fluctuation attribute as a feature vector, taking the influence value corresponding to the fluctuation attribute as a target value, and training a preset algorithm model to obtain a functional relation between the fluctuation attribute and the influence value;
and determining an initial estimation function corresponding to the fluctuation attribute according to the functional relation.
7. The method of claim 6, wherein the fluctuating attributes comprise vehicle condition information, the vehicle condition information comprising: before training a preset algorithm model by using the value of the fluctuation attribute as a feature vector and the influence value corresponding to the fluctuation attribute as a target value, the accident information and the maintenance information of the vehicle further include:
obtaining vehicles with the same vehicle age information in the sample vehicle estimation data to obtain a data set of a plurality of vehicles with the same vehicle age information;
clustering the accident information of the vehicles in the data sets of the vehicles with the same vehicle age information to obtain a plurality of groups of similar accident information sets;
clustering the maintenance information of the vehicles in the accident information set to obtain a plurality of groups of similar maintenance information sets;
and acquiring the influence value of the vehicle in each group of the maintenance information sets.
8. The method according to claim 7, wherein the training of the preset algorithm model by using the value of the fluctuation attribute as a feature vector and the influence value corresponding to the fluctuation attribute as a target value comprises:
for each maintenance information set, taking the vehicle age information, the accident information and the maintenance information as characteristic vectors, taking the influence value of the vehicle in the maintenance information set as a target value, and training a preset algorithm model to obtain a functional relation between the vehicle age information and the influence value corresponding to the maintenance information set;
the function relationship is a piecewise function which takes the vehicle age information as an independent variable and the influence value as a dependent variable, each subsection of the piecewise function is a constant function, and the value of the constant function is reduced along with the increase of the vehicle age information.
9. The method of claim 6, wherein the fluctuation attribute includes mileage information, and for each of the mileage information values, a functional relationship between age information corresponding to the mileage information value and the influence value is: and a linear function with the vehicle age information as an independent variable and the influence value as a dependent variable, wherein the slope of the linear function is a negative number.
10. The method of claim 6, wherein the fluctuation attribute is month information of the vehicle transaction, and the functional relationship between the month information and the impact value is: and a broken line type function which takes the month as an independent variable and the influence value as a dependent variable.
11. The method according to claim 2, wherein the attribute information includes at least model information, optional installation location information, license plate information, usage information, and external environment information of the vehicle to be evaluated.
12. The method of claim 11, wherein the license plate information in the stable attribute comprises boarding information, license plate attribute information; the use information in the stable attribute comprises use property information and passing times information, and the external environment information in the stable attribute comprises country preferential policy information.
13. The method of claim 11, wherein the usage information in the volatility attributes comprises age information, condition information, mileage information, claim information, and the external policy information in the volatility attributes comprises month information for the vehicle transaction.
14. A vehicle estimation device, characterized by comprising:
the attribute information acquisition module is used for acquiring attribute information of the vehicle to be evaluated;
the attribute type determining module is used for determining the attribute type corresponding to each attribute information, and the attribute type column determines the influence degree of the attribute information on the vehicle valuation along with the time;
a data input module for inputting the attribute information and the attribute category into a target vehicle valuation model; the estimation function in the target vehicle estimation model comprises sub estimation functions corresponding to the attribute information and weights corresponding to the attribute types; the target vehicle estimation model is constructed according to vehicle estimation data in a historical time period;
and the estimation module is used for obtaining an estimation result of the vehicle to be estimated according to the attribute information and the estimation function.
15. An electronic device comprising a processor and a memory having stored therein at least one instruction, at least one program, a set of codes, or a set of instructions, which is loaded and executed by the processor to implement the vehicle estimation method according to any one of claims 1 to 13.
16. A computer readable storage medium, characterized in that at least one instruction, at least one program, a set of codes, or a set of instructions is stored in the storage medium, which is loaded and executed by a processor to implement the vehicle estimation method according to any one of claims 1-13.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210719439.4A CN115310256A (en) | 2022-06-23 | 2022-06-23 | Vehicle estimation method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210719439.4A CN115310256A (en) | 2022-06-23 | 2022-06-23 | Vehicle estimation method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115310256A true CN115310256A (en) | 2022-11-08 |
Family
ID=83855171
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210719439.4A Pending CN115310256A (en) | 2022-06-23 | 2022-06-23 | Vehicle estimation method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115310256A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115880565A (en) * | 2022-12-06 | 2023-03-31 | 江苏凤火数字科技有限公司 | Neural network-based scraped vehicle identification method and system |
| CN117670382A (en) * | 2023-12-05 | 2024-03-08 | 广州穗圣信息科技有限公司 | Method and system for carrying out secondary handcart estimation by utilizing big data |
-
2022
- 2022-06-23 CN CN202210719439.4A patent/CN115310256A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115880565A (en) * | 2022-12-06 | 2023-03-31 | 江苏凤火数字科技有限公司 | Neural network-based scraped vehicle identification method and system |
| CN115880565B (en) * | 2022-12-06 | 2023-09-05 | 江苏凤火数字科技有限公司 | Neural network-based scraped vehicle identification method and system |
| CN117670382A (en) * | 2023-12-05 | 2024-03-08 | 广州穗圣信息科技有限公司 | Method and system for carrying out secondary handcart estimation by utilizing big data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10366435B2 (en) | Vehicle data system for rules based determination and real-time distribution of enhanced vehicle data in an online networked environment | |
| US9836714B2 (en) | Systems and methods for determining costs of vehicle repairs and times to major repairs | |
| US20100179861A1 (en) | System and method for assessing and managing objects | |
| CN115310256A (en) | Vehicle estimation method and device | |
| US20170316459A1 (en) | Data system for adaptive incentive allocation in an online networked environment | |
| US20140279263A1 (en) | Systems and methods for providing product recommendations | |
| US20170032400A1 (en) | Vehicle data system for distribution of vehicle data in an online networked environment | |
| US20150254719A1 (en) | Prediction of Vehicle Transactions and Targeted Advertising Using Vehicle Telematics | |
| US8620763B2 (en) | System, method and computer program product for demand-weighted selection of sales outlets | |
| Tsai et al. | Customer segmentation issues and strategies for an automobile dealership with two clustering techniques | |
| US10878435B2 (en) | Method and system for presenting information for a geographically eligible set of automobile dealerships ranked based on likelihood scores | |
| CA2837338A1 (en) | System and method for analysis and presentation of used vehicle pricing data | |
| CN108154275A (en) | Automobile residual value prediction model and Forecasting Methodology based on big data | |
| US20150332284A1 (en) | System and method for determining service intervals based on transaction data | |
| CA2829557A1 (en) | Wholesale/trade-in pricing system, method and computer program product therefor | |
| US10410227B2 (en) | System, method, and computer program for forecasting residual values of a durable good over time | |
| CA2837449A1 (en) | System, method and computer program product for geo-specific vehicle pricing | |
| US20200027141A1 (en) | System and method for analysis and presentation of used vehicle pricing data | |
| Robinson et al. | Issues related to field reliability and warranty data | |
| US20220335359A1 (en) | System and method for comparing enterprise performance using industry consumer data in a network of distributed computer systems | |
| Li et al. | The cost of greening stimulus: A dynamic analysis of vehicle scrappage programs | |
| US11915275B2 (en) | Systems and methods for estimating asset resale value | |
| CN111047438A (en) | Data processing method, device and computer readable storage medium | |
| US20230097392A1 (en) | Evaluation of a vehicle service for a vehicle based on information associated with a user of the vehicle | |
| Aguirregabiria et al. | Decentralized decision-making in retail chains: Evidence from inventory management |
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 |