CN113657945A - User value prediction method, device, electronic equipment and computer storage medium - Google Patents
User value prediction method, device, electronic equipment and computer storage medium Download PDFInfo
- Publication number
- CN113657945A CN113657945A CN202110995564.3A CN202110995564A CN113657945A CN 113657945 A CN113657945 A CN 113657945A CN 202110995564 A CN202110995564 A CN 202110995564A CN 113657945 A CN113657945 A CN 113657945A
- Authority
- CN
- China
- Prior art keywords
- data
- historical
- user
- time period
- prediction
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 60
- 238000012549 training Methods 0.000 claims abstract description 43
- 238000012545 processing Methods 0.000 claims abstract description 19
- 238000012795 verification Methods 0.000 claims description 13
- 238000010801 machine learning Methods 0.000 claims description 9
- 238000000605 extraction Methods 0.000 claims description 7
- 238000010276 construction Methods 0.000 claims description 4
- 238000005457 optimization Methods 0.000 claims description 3
- 238000004422 calculation algorithm Methods 0.000 description 6
- 230000002860 competitive effect Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 2
- 238000003066 decision tree Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000006399 behavior Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000002790 cross-validation Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000004806 packaging method and process Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- 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
- G06Q30/0203—Market surveys; Market polls
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Finance (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Economics (AREA)
- Game Theory and Decision Science (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a user value prediction method, a user value prediction device, electronic equipment and a computer storage medium, and user data are acquired; generating a feature data set based on feature data extracted from the user data; and taking the characteristic data set as the input of a pre-constructed data prediction model, and processing the characteristic data set based on the pre-constructed data prediction model to obtain the user value in a user preset time period, wherein the pre-constructed data prediction model is obtained by training the historical data of each user. In the scheme, the feature data in the acquired user data are extracted to generate a corresponding feature data set. And processing the characteristic data set by using a pre-constructed data prediction model so as to predict the user value in a user preset time period. By the method, the user value of each user can be known, so that marketing and user operation departments can make targeted marketing strategies, and the stickiness of the users to the fund companies is further improved.
Description
Technical Field
The present invention relates to the field of data processing technologies, and in particular, to a user value prediction method, an apparatus, an electronic device, and a computer storage medium.
Background
With the advent of the big data age, aiming marketing strategies are made for the marketing and user operation departments to know the value of users, so that the stickiness of users to fund companies is improved.
At present, a Customer life cycle Value (CLV) processes input multiple purchase amounts and purchase times of a user to calculate a user Value. Because the CLV model only considers the procurement behavior and can be effectively used on the premise that the consumption amount of the user accords with the gamma-gamma distribution data distribution hypothesis, the user value calculated by adopting the CLV model is inaccurate.
Disclosure of Invention
In view of this, embodiments of the present invention provide a user value prediction method, an apparatus, an electronic device, and a computer storage medium, so as to solve the problem in the prior art that a user value calculated by a CLV model is inaccurate.
In order to achieve the above purpose, the embodiments of the present invention provide the following technical solutions:
the first aspect of the embodiments of the present invention shows a user value prediction method, where the method includes:
acquiring user data, wherein the user data at least comprises transaction information, browsing data and user attribute data;
generating a feature data set based on feature data extracted from the user data;
and taking the characteristic data set as the input of a pre-constructed data prediction model, and processing the characteristic data set based on the pre-constructed data prediction model to obtain the user value in the user preset time period, wherein the pre-constructed data prediction model is obtained by training the historical data of each user.
Optionally, the process of obtaining the data prediction model by training using the historical data and the prediction data of each user includes:
acquiring historical data of each user in a historical time period;
determining historical data and historical prediction data before a first historical moment based on the historical data of each user in the historical time period;
extracting characteristic data in historical data before the first historical moment;
dividing the feature data and historical prediction data into a training set and a verification set;
training a universal machine learning model based on the training set to obtain an initial data prediction model;
and optimizing the initial data prediction model on the verification set to obtain a trained data prediction model.
Optionally, the determining historical data and historical prediction data before the first historical time based on the historical data of each user in the historical time period includes:
acquiring the held asset share data of each user in the historical data within a first time period, wherein the first time period is the time between a first historical moment and a second historical moment;
for each user, calculating asset share data in a first time period to obtain historical prediction data, wherein the historical prediction data comprises an asset share daily average value corresponding to each user;
and determining historical data before the first historical time based on the historical data.
Optionally, the determining historical data and historical prediction data before the first historical time based on the historical data of each user in the historical time period includes:
acquiring the held asset share data of each user in the historical data within a first time period, wherein the first time period is the time between a first historical moment and a second historical moment;
for each user, calculating the asset share data in a first time period to obtain an asset share daily average value corresponding to each user;
for each user, calculating based on the daily average value of the asset shares and the asset shares at the first historical moment to obtain historical prediction data, wherein the historical prediction data comprises an asset share proportion value corresponding to each user;
and determining historical data before the first historical time based on the historical data.
A second aspect of the embodiments of the present invention shows a user value prediction apparatus, including:
the system comprises an acquisition unit, a processing unit and a processing unit, wherein the acquisition unit is used for acquiring user data, and the user data at least comprises transaction information, browsing data and user attribute data;
a generation unit configured to generate a feature data set based on feature data extracted from the user data;
and the data prediction model is used for taking the characteristic data set as the input of a pre-constructed data prediction model, processing the characteristic data set based on the pre-constructed data prediction model and obtaining the user value in the user preset time period, wherein the pre-constructed data prediction model is constructed by the construction unit.
Optionally, the building unit includes:
the acquisition module is used for acquiring historical data of each user in a historical time period;
the determining module is used for determining historical data and historical prediction data before a first historical moment based on the historical data of each user in the historical time period;
the extraction module is used for extracting characteristic data in the historical data before the first historical moment;
the dividing module is used for dividing the characteristic data and the historical prediction data into a training set and a verification set;
the training module is used for training a universal machine learning model based on the training set to obtain an initial data prediction model;
and the optimization module is used for optimizing the initial data prediction model on the verification set to obtain a trained data prediction model.
Optionally, the extracting module is specifically configured to: acquiring the held asset share data of each user in the historical data within a first time period, wherein the first time period is the time between a first historical moment and a second historical moment; for each user, calculating asset share data in a first time period to obtain historical prediction data, wherein the historical prediction data comprises an asset share daily average value corresponding to each user; and determining historical data before the first historical time based on the historical data.
Optionally, the extracting module is specifically configured to: acquiring the held asset share data of each user in the historical data within a first time period, wherein the first time period is the time between a first historical moment and a second historical moment; for each user, calculating the asset share data in a first time period to obtain an asset share daily average value corresponding to each user; for each user, calculating based on the daily average value of the asset shares and the asset shares at the first historical moment to obtain historical prediction data, wherein the historical prediction data comprises an asset share proportion value corresponding to each user; and determining historical data before the first historical time based on the historical data.
A third aspect of the embodiments of the present invention shows an electronic device, where the electronic device is configured to run a program, where the program executes the user value prediction method shown in the first aspect of the embodiments of the present invention when running.
A fourth aspect of the embodiments of the present invention shows a computer storage medium, where the storage medium includes a storage program, and when the program runs, a device in which the storage medium is located is controlled to execute the user value prediction method shown in the first aspect of the embodiments of the present invention.
Based on the user value prediction method, the user value prediction device, the electronic equipment and the computer storage medium, provided by the embodiment of the invention, the method comprises the following steps: acquiring user data, wherein the user data at least comprises transaction information, browsing data and user attribute data; generating a feature data set based on feature data extracted from the user data; and taking the characteristic data set as the input of a pre-constructed data prediction model, and processing the characteristic data set based on the pre-constructed data prediction model to obtain the user value in a user preset time period, wherein the pre-constructed data prediction model is obtained by training the historical data of each user. In the embodiment of the invention, the user data is preprocessed to extract the feature data in the user data and generate the corresponding feature data set. And processing the characteristic data set by using a pre-constructed data prediction model so as to predict the user value in a user preset time period. By the method, the user value of each user can be known, so that marketing and user operation departments can make targeted marketing strategies, and the stickiness of the users to the fund companies is further improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
Fig. 1 is a schematic flow chart of a user value prediction method according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a training data prediction model according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of a user value prediction apparatus 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 derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In this application, 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 phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
In the embodiment of the invention, the user data is preprocessed to extract the feature data in the user data and generate the corresponding feature data set. And processing the characteristic data set by using a pre-constructed data prediction model so as to predict the user value in a user preset time period. By the method, the user value of each user can be known, so that marketing and user operation departments can make targeted marketing strategies, and the stickiness of the users to the fund companies is further improved.
Referring to fig. 1, a schematic flow chart of a user value prediction method provided in an embodiment of the present invention is shown, where the method includes:
s101: user data is acquired.
In step S101, the user data includes at least transaction information, browsing data, and user attribute data.
Specific contents of S101: when a user identification number input by an operator is detected, traversing the database, and acquiring transaction information, browsing data, user identity data and the like of a user corresponding to the user identification number before the current time.
It is further noted that the user data number includes historical property information, user profit and loss information, and market and competitive information.
Wherein the user identification number is used for uniquely identifying the user.
The user indicated in the embodiment of the present invention is a user of a fund company.
S102: a feature data set is generated based on feature data extracted from the user data.
Specific contents of S102: the method comprises the steps of extracting transaction information, browsing data, user attribute data, historical held asset information, user profit and loss information and characteristic data in market and competitive product information, and packaging the extracted characteristic data to generate a characteristic data set.
In the embodiment of the invention, the transaction information, the browsing data, the user attribute data, the historical property information, the user profit and loss information and the characteristic data in the market and competitive product information are extracted, as shown in the table (1).
Watch (1)
Extracting information such as age, gender, region, user type, transaction channel and the like in the user attribute information; extracting information such as latest purchase amount, average monthly stroke number, purchase redemption amount and the like in the transaction information; extracting information such as a balance mean value, a maximum asset, a variation coefficient and the like in the historical asset information; extracting total profit and loss amount in the profit and loss information of the user, the profit and loss amount in different recent periods and the like; extracting the difference value between the redemption time point yield and the balance treasure value, the Shanghai depth of 300 and the like in the market and competitive product information; and extracting information such as login time, reading data, collecting data, clicking data, marketing activities and the like in the browsing data.
It should be noted that the user types are divided by investing funds of the users; the transaction channels refer to different applications used to purchase funds; the latest subscription amount may be a subscription amount within the last week; the amount of profit and loss in different recent periods can be the amount of profit and loss in different recent periods in the last week;
s103: and taking the characteristic data set as the input of a pre-constructed data prediction model, and processing the characteristic data based on the pre-constructed data prediction model to obtain the user value in the user preset time period.
In step S103, the pre-constructed data prediction model is trained using the historical data of each user.
It should be noted that, the process of training and obtaining the data prediction model by using the historical data of each user includes the following steps:
correspondingly, the embodiment of the application also discloses an architecture schematic diagram of the training data prediction model, as shown in fig. 2.
S11: historical data of each user in a historical time period is acquired.
In step S11, the historical data refers to transaction information, browsing data, user attribute data, historical property information, user profit and loss information, market and competitive goods information, etc. of each user in the historical time period.
It should be noted that the historical time period is preset by the technician, and may be set to be the past year, for example.
The historical time period includes a first time period, that is, the time corresponding to the historical time period is longer than the time of the first time period.
S12: and determining historical data and historical prediction data before the first historical time based on the historical data of each user in the historical time period.
In the embodiment of the invention, the historical prediction data is the daily average value of the asset share corresponding to each user, or the proportional value of the asset share corresponding to each user; historical data before the first historical time and the daily average value of the asset share corresponding to each user can be used as a data set of the training data prediction model, and historical data before the first historical time and the proportional value of the asset share corresponding to each user can also be used as a data set of the training data prediction model. The following description will be made separately.
In the first embodiment, the step S12 of determining the history data and the history prediction data before the first history time based on the history data of each user in the history time period includes the following steps:
s21: the held asset share data for each user in the historical data for the first time period is obtained.
In step S21, the first period of time is the time between the first history time and the second history time.
The specific procedure of step S21 is to acquire change data of the actual share of the owned asset of each user in the time period from the first history time to the second history time.
For example: suppose the first history time is T0The second history time is TeAt this time, [ T ] needs to be acquired0,Te]The actual holding asset share of the time period users has changed data.
It should be noted that the asset share data refers to the recorded asset share data held at each time.
S22: and calculating the change data of the actual held asset share in the first time period aiming at each user to obtain historical prediction data.
In step S22, the historical forecast data includes the daily average of the share of the assets corresponding to each user.
Specific contents of S22: substituting the change data x of the actual held asset share in the first time period into a formula (1) for each user to calculate to obtain an asset share daily average value v corresponding to each userav。
Formula (1):
where x is the change data of the actual share of the held asset over the first time period and τ (t) represents the time worth factor.
τ (t) refers to a time factor function related to profit, that is, the magnitude of τ (t) is related to interest rate.
The size of τ (t) is set by the skilled person according to the actual situation, and in particular τ (t) can be set to be a constant of 1 without considering the influence of interest rate.
Such as: the capital share is 100 shares and the investment is 10 months, the capital share is 1000 yuan and the investment is 1 month corresponding to vavAre the same.
S23: based on the historical data, historical data prior to the first historical time is determined.
Specific contents of S23:obtaining a first historical time T of each user from historical data0Previous historical data such as historical held asset information, transaction information, user attribute data, and user profit and loss information.
In the second embodiment, the process of determining the history data and the history prediction data before the first history time based on the history data of each user in the history time period in the step S12 includes the following steps:
s31: the held asset share data for each user in the historical data for the first time period is obtained.
In step S31, the first period of time is the time between the first history time and the second history time.
S32: and calculating the actual held asset share in the first time period aiming at each user to obtain the daily average of the asset share corresponding to each user.
It should be noted that the specific implementation contents of step S31 and step S32 are the same as those of step S21 and step S22 described above.
It should be noted that the asset share data refers to data corresponding to the change of the asset share of the user at each time in the first time period, that is, the asset share data includes the asset share at the first historical time.
S33: and calculating the historical prediction data based on the daily average value of the asset shares and the asset shares at the first historical moment aiming at each user to obtain the historical prediction data, wherein the historical prediction data comprises the asset share proportion value corresponding to each user.
Specific contents of S33: for each user, the daily average v of the share of the assetsavAnd asset share v at a first historical time0And substituting the obtained value into the formula (2) for calculation to obtain the asset share proportion value R corresponding to each user.
Formula (2):
wherein v isavIs calculated by the formula (1)The daily average of the share of assets of (1); v. of0Is the asset share at the first historical time.
S34: based on the historical data, historical data prior to the first historical time is determined.
It should be noted that the specific implementation process of step S34 is the same as the specific implementation process of step S23, and can be referred to each other.
S13: and extracting characteristic data in the historical data before the first historical time.
Specific contents of S13: and extracting transaction information, browsing data, user attribute data, history held asset information, user profit and loss information and characteristic data in market and competitive product information before the first historical moment so as to create a sample point for each user.
It should be noted that the specific implementation process of step S13 is the same as the specific implementation process of step S102, and reference may be made to each other.
S14: the feature data and historical prediction data are partitioned into a training set and a validation set.
S15: and training the universal machine learning model based on the training set to obtain an initial data prediction model.
S16: and optimizing the initial data prediction model on the verification set to obtain the trained data prediction model.
It should be noted that the general Machine learning model may be a Light Gradient Boosting Machine (LightGBM) algorithm model.
In the specific implementation process from step S14 to step S16, feature data and historical prediction data corresponding to the feature data are divided into K groups (K-Fold) according to a preset proportion, each subset data is made into a verification set, and the rest K-1 groups of subset data are used as training sets. And configuring network parameters of the initial data prediction model based on the LightGBM algorithm. Firstly, training network parameters based on a training set, constructing an initial data prediction model by using the network parameters, and then evaluating a verification set by using the initial data prediction model to obtain an evaluation result. The cross-validation is repeated k times, where k may be set to 5.
The initial data prediction model corresponding to each training set needs to be evaluated once, the average value of k evaluation results is taken to optimize network parameters, the generalization error can be reduced, and the data prediction model is constructed based on the optimized network parameters.
It should be noted that the preset ratio can be set to 4:1, which can be set according to practical situations, and the application is not limited.
The feature data and the historical prediction data of the same user are in the same data set, such as: both the feature data and the historical prediction data for user a are in the training set.
The LightGBM algorithm model is an algorithm model which is fast, efficient, low in memory occupation, high in accuracy and capable of supporting parallel and large-scale data processing.
Further, it should be noted that, not only the machine learning model shown above, but also a support vector machine, a decision tree, a neighbor algorithm (KNN), a Random Forest format, a Gradient Boosting decision tree model (XGBoost), a multilayer perceptron MLP, a Deep Neural Network (DNN), a recurrent Neural network RNN, a Gradient descending tree (GBDT), a convolutional Neural network CNN, or other machine learning models may be used for training set training.
Based on this, the historical prediction data of the training data prediction model is different, and when the user value is predicted by using the data prediction model, the description mode of the obtained user value is also different.
The user value may be represented by the daily average of the holding shares, [ T [ ]0，Te]The user value may be defined by mathematical statistics such as the total of the share of the assets held in between, the maximum daily share value, or the minimum daily share value, or may be defined by such statistics relative to a certain point in time (e.g., T)0) The user value may be defined in different ways, such as a combined operation value of the unique asset shares. The combined operation value may be a proportional value of the statistic with respect to the unique asset share at a certain time point, or a rate of change of the statistic with respect to the unique asset share at a certain time point, or the like.
Wherein the user value includes at least a predicted asset share and a predicted growth rate.
The predicted growth rate refers to a relative change amount, i.e., a value of a ratio of the calculated future value to the current value.
In the embodiment of the present invention, when historical data before the first historical time and the daily average of the share of the assets corresponding to each user are used as a data set of a training data prediction model, the process of specifically implementing step S103 includes: inputting the feature data set, i.e. the test set, into the framework of the trained data prediction model shown in fig. 2, so that the data prediction model processes the feature data set to obtain the predicted asset share in the preset time period.
The data prediction model described above may predict not only the share of assets, but also the user value such as the owned asset value, the management fee of the user fund asset, the net value of the user asset, and the like, that is, the user value may be a time value such as the owned asset value, the management fee, and the like.
In the embodiment of the present invention, when historical data before the first historical time and an asset share ratio value corresponding to each user are used as a data set of a training data prediction model, a process of specifically implementing step S103 includes: inputting the feature data set, i.e. the test set, into the architecture of the trained data prediction model in fig. 2, so that the data prediction model processes the feature data set to obtain the predicted growth ratio in the preset time period.
It should be noted that the predicted increase rate may be an asset change rate or a fund management fee change rate, or the like.
It should be noted that the preset time is set by a technician in advance according to actual conditions, and may be set to a period of 1 month, half year, one year, and the like in the future. The preset time is as long as the first time period. That is, to predict the user value of a user for one month in the future, the first time period when training the data is also 1 month.
In the embodiment of the invention, the user data is preprocessed to extract the feature data in the user data and generate the corresponding feature data set. And processing the characteristic data set by using a pre-constructed data prediction model so as to predict the user value in a user preset time period. By the method, the user value of each user can be known, so that marketing and user operation departments can make targeted marketing strategies, and the stickiness of the users to the fund companies is further improved.
Corresponding to the user value prediction method shown in the above embodiment of the present invention, the embodiment of the present invention also discloses a schematic structural diagram of a user value prediction apparatus, as shown in fig. 3, the apparatus includes:
the acquiring unit 301 is configured to acquire user data, where the user data at least includes transaction information, browsing data, and user attribute data.
A generating unit 302 for generating a feature data set based on feature data extracted from the user data.
And the data prediction model 303 is configured to take the feature data set as an input of a pre-constructed data prediction model, and process the feature data set based on the pre-constructed data prediction model to obtain a user value in a user preset time period, where the pre-constructed data prediction model is constructed by the construction unit 304.
It should be noted that, the specific principle and the execution process of each unit in the user value prediction apparatus disclosed in the embodiment of the present application are the same as those of the user value prediction method described in the embodiment of the present application, and reference may be made to corresponding parts in the data processing method disclosed in the embodiment of the present application, which are not described herein again.
In the embodiment of the invention, the user data is preprocessed to extract the feature data in the user data and generate the corresponding feature data set. And processing the characteristic data set by using a pre-constructed data prediction model so as to predict the user value in a user preset time period. By the method, the user value of each user can be known, so that marketing and user operation departments can make targeted marketing strategies, and the stickiness of the users to the fund companies is further improved.
Based on the user value prediction apparatus shown in the above embodiment of the present invention, the construction unit 304 includes:
the acquisition module is used for acquiring historical data of each user in a historical time period;
the determining module is used for determining historical data and historical prediction data before a first historical moment based on the historical data of each user in a historical time period;
the extraction module is used for extracting characteristic data in the historical data before the first historical moment;
the dividing module is used for dividing the characteristic data and the historical prediction data into a training set and a verification set;
the training module is used for training a universal machine learning model based on a training set to obtain an initial data prediction model;
and the optimization module is used for optimizing the initial data prediction model on the verification set to obtain the trained data prediction model.
Optionally, the extraction module is specifically configured to: acquiring the held asset share data of each user in historical data within a first time period, wherein the first time period is the time between a first historical moment and a second historical moment; for each user, calculating the asset share data in a first time period to obtain historical prediction data, wherein the historical prediction data comprises an asset share daily average value corresponding to each user; based on the historical data, historical data prior to the first historical time is determined.
Optionally, the extraction module is specifically configured to: acquiring the held asset share data of each user in historical data within a first time period, wherein the first time period is the time between a first historical moment and a second historical moment; for each user, calculating the asset share data in a first time period to obtain an asset share daily average value corresponding to each user; calculating for each user based on the daily average value of the asset shares and the asset shares at the first historical moment to obtain historical prediction data, wherein the historical prediction data comprises an asset share proportion value corresponding to each user; based on the historical data, historical data prior to the first historical time is determined.
In the embodiment of the invention, the user data is preprocessed to extract the feature data in the user data and generate the corresponding feature data set. And processing the characteristic data set by using a pre-constructed data prediction model so as to predict the user value in a user preset time period. By the method, the user value of each user can be known, so that marketing and user operation departments can make targeted marketing strategies, and the stickiness of the users to the fund companies is further improved.
The embodiment of the invention also discloses electronic equipment, which is used for operating the database storage process, wherein the user value prediction method disclosed in the figure 1 is executed when the database storage process is operated.
The embodiment of the invention also discloses a computer storage medium, which comprises a storage database storage process, wherein when the storage database storage process runs, the equipment where the storage medium is located is controlled to execute the user value prediction method disclosed in the figure 1.
In the context of this disclosure, a computer storage medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, the system or system embodiments are substantially similar to the method embodiments and therefore are described in a relatively simple manner, and reference may be made to some of the descriptions of the method embodiments for related points. The above-described system and system embodiments are only illustrative, wherein the units described as separate parts may or may not be physically separate, and the parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Those of skill would further appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various illustrative components and steps have been described above generally in terms of their functionality in order to clearly illustrate this interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. A method for predicting user value, the method comprising:
acquiring user data, wherein the user data at least comprises transaction information, browsing data and user attribute data;
generating a feature data set based on feature data extracted from the user data;
and taking the characteristic data set as the input of a pre-constructed data prediction model, and processing the characteristic data set based on the pre-constructed data prediction model to obtain the user value in the user preset time period, wherein the pre-constructed data prediction model is obtained by training the historical data of each user.
2. The method of claim 1, wherein the training with the historical data and the predictive data of each user to obtain a data prediction model comprises:
acquiring historical data of each user in a historical time period;
determining historical data and historical prediction data before a first historical moment based on the historical data of each user in the historical time period;
extracting characteristic data in historical data before the first historical moment;
dividing the feature data and historical prediction data into a training set and a verification set;
training a universal machine learning model based on the training set to obtain an initial data prediction model;
and optimizing the initial data prediction model on the verification set to obtain a trained data prediction model.
3. The method of claim 2, wherein determining historical data and historical predicted data prior to a first historical time based on the historical data for each user in the historical time period comprises:
acquiring the held asset share data of each user in the historical data within a first time period, wherein the first time period is the time between a first historical moment and a second historical moment;
for each user, calculating asset share data in a first time period to obtain historical prediction data, wherein the historical prediction data comprises an asset share daily average value corresponding to each user;
and determining historical data before the first historical time based on the historical data.
4. The method of claim 2, wherein determining historical data and historical predicted data prior to a first historical time based on the historical data for each user in the historical time period comprises:
acquiring the held asset share data of each user in the historical data within a first time period, wherein the first time period is the time between a first historical moment and a second historical moment;
for each user, calculating the asset share data in a first time period to obtain an asset share daily average value corresponding to each user;
for each user, calculating based on the daily average value of the asset shares and the asset shares at the first historical moment to obtain historical prediction data, wherein the historical prediction data comprises an asset share proportion value corresponding to each user;
and determining historical data before the first historical time based on the historical data.
5. An apparatus for predicting user value, the apparatus comprising:
the system comprises an acquisition unit, a processing unit and a processing unit, wherein the acquisition unit is used for acquiring user data, and the user data at least comprises transaction information, browsing data and user attribute data;
a generation unit configured to generate a feature data set based on feature data extracted from the user data;
and the data prediction model is used for taking the characteristic data set as the input of a pre-constructed data prediction model, processing the characteristic data set based on the pre-constructed data prediction model and obtaining the user value in the user preset time period, wherein the pre-constructed data prediction model is constructed by the construction unit.
6. The apparatus of claim 5, wherein the building unit comprises:
the acquisition module is used for acquiring historical data of each user in a historical time period;
the determining module is used for determining historical data and historical prediction data before a first historical moment based on the historical data of each user in the historical time period;
the extraction module is used for extracting characteristic data in the historical data before the first historical moment;
the dividing module is used for dividing the characteristic data and the historical prediction data into a training set and a verification set;
the training module is used for training a universal machine learning model based on the training set to obtain an initial data prediction model;
and the optimization module is used for optimizing the initial data prediction model on the verification set to obtain a trained data prediction model.
7. The apparatus according to claim 6, wherein the extraction module is specifically configured to: acquiring the held asset share data of each user in the historical data within a first time period, wherein the first time period is the time between a first historical moment and a second historical moment; for each user, calculating asset share data in a first time period to obtain historical prediction data, wherein the historical prediction data comprises an asset share daily average value corresponding to each user; and determining historical data before the first historical time based on the historical data.
8. The apparatus according to claim 6, wherein the extraction module is specifically configured to: acquiring the held asset share data of each user in the historical data within a first time period, wherein the first time period is the time between a first historical moment and a second historical moment; for each user, calculating the asset share data in a first time period to obtain an asset share daily average value corresponding to each user; for each user, calculating based on the daily average value of the asset shares and the asset shares at the first historical moment to obtain historical prediction data, wherein the historical prediction data comprises an asset share proportion value corresponding to each user; and determining historical data before the first historical time based on the historical data.
9. An electronic device, wherein the electronic device is configured to run a program, wherein the program when running performs the user value prediction method according to any one of claims 1-4.
10. A computer storage medium, characterized in that the storage medium comprises a stored program, wherein the program, when executed, controls an apparatus in which the storage medium is located to perform the user value prediction method according to any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110995564.3A CN113657945A (en) | 2021-08-27 | 2021-08-27 | User value prediction method, device, electronic equipment and computer storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110995564.3A CN113657945A (en) | 2021-08-27 | 2021-08-27 | User value prediction method, device, electronic equipment and computer storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113657945A true CN113657945A (en) | 2021-11-16 |
Family
ID=78493097
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110995564.3A Pending CN113657945A (en) | 2021-08-27 | 2021-08-27 | User value prediction method, device, electronic equipment and computer storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113657945A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117688343A (en) * | 2024-02-04 | 2024-03-12 | 成都帆点创想科技有限公司 | LTV prediction method and system for multi-task learning LSTM-Attention framework |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070112618A1 (en) * | 2005-11-09 | 2007-05-17 | Generation 5 Mathematical Technologies Inc. | Systems and methods for automatic generation of information |
| CN109919684A (en) * | 2019-03-18 | 2019-06-21 | 上海盛付通电子支付服务有限公司 | For generating method, electronic equipment and the computer readable storage medium of information prediction model |
| CN110110012A (en) * | 2019-04-23 | 2019-08-09 | 上海淇玥信息技术有限公司 | User's expectancy appraisal procedure, device, electronic equipment and readable medium |
| CN110415036A (en) * | 2019-07-30 | 2019-11-05 | 深圳市珍爱捷云信息技术有限公司 | Determination method, apparatus, computer equipment and the storage medium of user gradation |
| CN111695719A (en) * | 2020-04-20 | 2020-09-22 | 清华大学 | User value prediction method and system |
| CN112116159A (en) * | 2020-09-21 | 2020-12-22 | 贝壳技术有限公司 | Information interaction method and device, computer readable storage medium and electronic equipment |
| CN112990989A (en) * | 2021-05-17 | 2021-06-18 | 太平金融科技服务(上海)有限公司深圳分公司 | Value prediction model input data generation method, device, equipment and medium |
| CN113205367A (en) * | 2021-05-24 | 2021-08-03 | 上海钧正网络科技有限公司 | User data processing method and device, electronic equipment and storage medium |
-
2021
- 2021-08-27 CN CN202110995564.3A patent/CN113657945A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070112618A1 (en) * | 2005-11-09 | 2007-05-17 | Generation 5 Mathematical Technologies Inc. | Systems and methods for automatic generation of information |
| CN109919684A (en) * | 2019-03-18 | 2019-06-21 | 上海盛付通电子支付服务有限公司 | For generating method, electronic equipment and the computer readable storage medium of information prediction model |
| CN110110012A (en) * | 2019-04-23 | 2019-08-09 | 上海淇玥信息技术有限公司 | User's expectancy appraisal procedure, device, electronic equipment and readable medium |
| CN110415036A (en) * | 2019-07-30 | 2019-11-05 | 深圳市珍爱捷云信息技术有限公司 | Determination method, apparatus, computer equipment and the storage medium of user gradation |
| CN111695719A (en) * | 2020-04-20 | 2020-09-22 | 清华大学 | User value prediction method and system |
| CN112116159A (en) * | 2020-09-21 | 2020-12-22 | 贝壳技术有限公司 | Information interaction method and device, computer readable storage medium and electronic equipment |
| CN112990989A (en) * | 2021-05-17 | 2021-06-18 | 太平金融科技服务(上海)有限公司深圳分公司 | Value prediction model input data generation method, device, equipment and medium |
| CN113205367A (en) * | 2021-05-24 | 2021-08-03 | 上海钧正网络科技有限公司 | User data processing method and device, electronic equipment and storage medium |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117688343A (en) * | 2024-02-04 | 2024-03-12 | 成都帆点创想科技有限公司 | LTV prediction method and system for multi-task learning LSTM-Attention framework |
| CN117688343B (en) * | 2024-02-04 | 2024-05-03 | 成都帆点创想科技有限公司 | LTV prediction method and system for multi-task learning LSTM-Attention framework |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106485562B (en) | Commodity information recommendation method and system based on user historical behaviors | |
| WO2019001120A1 (en) | Method and system for processing dynamic pricing data of commodity | |
| US20110208701A1 (en) | Computer-Implemented Systems And Methods For Flexible Definition Of Time Intervals | |
| EP2506207A1 (en) | Computer system, computer-implemented method and computer program product for managing operations of a system | |
| EP3371764A1 (en) | Systems and methods for pricing optimization with competitive influence effects | |
| CN109583729B (en) | Data processing method and device for platform online model | |
| WO2020024718A1 (en) | Method and device for predicting foreign exchange transaction volume | |
| JP2021099718A (en) | Information processing device and program | |
| CN113837806A (en) | Product traffic prediction method, device, electronic equipment and storage medium | |
| JP5251217B2 (en) | Sales number prediction system, operation method of sales number prediction system, and sales number prediction program | |
| Hanslo et al. | Machine learning models to predict agile methodology adoption | |
| Stoean et al. | Evolutionary regressor selection in ARIMA model for stock price time series forecasting | |
| KR102311107B1 (en) | Customer exit prevention method that provides a solution to prevent customer from leaving the deep learning solution platform that automatically creates a deep learning model, perfomred by a system to avoid leaving the customer | |
| CN113657945A (en) | User value prediction method, device, electronic equipment and computer storage medium | |
| WO2019221844A1 (en) | Item-specific value optimization tool | |
| CN111667307B (en) | Method and device for predicting financial product sales volume | |
| CN107480703A (en) | Transaction fault detection method and device | |
| KR20160071899A (en) | Apparatus for evaluating a value of technique, method thereof, and storage media for storing a program evaluating avalue of technique | |
| Cetinkaya et al. | New product launch decisions with robust optimization | |
| CN116862561B (en) | Product heat analysis method and system based on convolutional neural network | |
| JP5826891B1 (en) | Prediction device, prediction method, and computer program | |
| CN113538055B (en) | Sales volume estimation method, device and processor for Amazon platform commodity | |
| US20230394512A1 (en) | Methods and systems for profit optimization | |
| Agarwal et al. | Architecture for Evaluating Customer Retention Strategies | |
| CN117611364A (en) | Input cost estimated deviation measuring and calculating method and system thereof |
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 |