CN110648153A - Change tendency prediction method and device, electronic equipment and storage medium - Google Patents
Change tendency prediction method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN110648153A CN110648153A CN201810595922.XA CN201810595922A CN110648153A CN 110648153 A CN110648153 A CN 110648153A CN 201810595922 A CN201810595922 A CN 201810595922A CN 110648153 A CN110648153 A CN 110648153A
- Authority
- CN
- China
- Prior art keywords
- user
- tendency
- neural network
- deep neural
- machine
- 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
- 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
- G06Q30/0202—Market predictions or forecasting for commercial activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
- H04W24/08—Testing, supervising or monitoring using real traffic
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Game Theory and Decision Science (AREA)
- Economics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Networks & Wireless Communication (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Signal Processing (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the invention provides a method and a device for predicting change tendency, electronic equipment and a storage medium, wherein the method comprises the following steps: acquiring terminal bill data of a user in a preset time period; acquiring characteristic data for predicting whether the user has the tendency of changing the machine or not according to the terminal bill data; and detecting whether the user has a tendency to change the machine or not according to the feature data and a deep neural network prediction model obtained by pre-training. The embodiment of the invention improves the accuracy of predicting whether the user has the tendency of changing the machine.
Description
Technical Field
The embodiment of the invention relates to the technical field of communication, in particular to a method and a device for predicting change tendency, electronic equipment and a storage medium.
Background
The mobile user change prediction model is a main means for an operator to find users with the tendency of changing mobile terminals, and the accuracy of the model is very important. At present, a mobile user machine change prediction model is mainly used for prediction based on shallow machine learning algorithms such as logistic regression, decision trees, random forests and the like or weights of artificially set features. However, in the context of large data, these shallow algorithms have difficulty achieving high accuracy in predicting problems. At this time, since the potential switch machine user base is huge, the lower switch machine prediction accuracy also reduces the success rate of terminal marketing of the operator.
In summary, in the prior art, when predicting whether a user has a tendency to change a machine, the prediction accuracy is low.
Disclosure of Invention
The embodiment of the invention provides a method and a device for predicting a machine changing tendency, electronic equipment and a storage medium, which are used for solving the problem of low prediction precision when predicting whether a user has the machine changing tendency in the prior art.
In view of the foregoing problems, in a first aspect, an embodiment of the present invention provides a method for predicting a tendency to change a machine, where the method includes:
acquiring terminal bill data of a user in a preset time period;
acquiring characteristic data for predicting whether the user has the tendency of changing the machine or not according to the terminal bill data;
and detecting whether the user has a tendency to change the machine or not according to the feature data and a deep neural network prediction model obtained by pre-training.
In a second aspect, an embodiment of the present invention provides a device for predicting a tendency to change a machine, where the device includes:
the first acquisition module is used for acquiring terminal bill data of a user in a preset time period;
the second acquisition module is used for acquiring characteristic data for predicting whether the user has the tendency of changing the machine or not according to the terminal bill data;
and the detection module is used for detecting whether the user has a tendency to change the machine or not according to the feature data and a deep neural network prediction model obtained by pre-training.
In a third aspect, an embodiment of the present invention provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, where the processor implements the steps of the method for predicting the tendency of a machine change when executing the computer program.
In a fourth aspect, an embodiment of the present invention provides a non-transitory computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the steps of the method for predicting the tendency of a machine change.
According to the method, the device, the electronic equipment and the storage medium for predicting the machine changing tendency of the user, the terminal bill data of the user in the preset time period is obtained, the characteristic data for predicting whether the user has the machine changing tendency is obtained according to the terminal bill data, and finally whether the user has the machine changing tendency is detected according to the characteristic data and the deep neural network prediction model obtained through pre-training, so that the accurate identification of whether the user has the machine changing tendency by using the deep neural network prediction model is realized, and the accuracy of predicting the machine changing tendency of the user is 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, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
FIG. 1 is a flow chart illustrating steps of a method for predicting machine change tendency in an embodiment of the present invention;
FIG. 2 is a block diagram of a device for predicting a tendency of a machine change in an embodiment of the present invention;
fig. 3 shows a block diagram of an electronic device in an embodiment of the invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, 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 some, but not all, embodiments of the present invention. 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.
As shown in fig. 1, a flowchart of steps of a method for predicting a tendency of a machine change in an embodiment of the present invention is shown, where the method includes the following steps:
step 101: and acquiring terminal bill data of the user in a preset time period.
In this step, specifically, the terminal bill data of the user in the preset time period is acquired, so that the acquired terminal bill data can be used as original data for predicting whether the user has a tendency to change the machine, and a basis is provided for predicting the tendency to change the machine.
Specifically, when terminal billing data of the user in a preset time period is obtained, the terminal billing data of the user in the preset time period after desensitization processing may be obtained from the charging center and the core network side, so as to protect the terminal number privacy of the user corresponding to the obtained terminal billing data.
Further, specifically, the preset time period may be three months. Of course, it should be noted herein that the value of the preset time period may be determined according to actual situations, and the specific value of the preset time period is not specifically limited herein.
In addition, the terminal billing data may specifically include Average Revenue Per User (ARPU), basic monthly rental, online time (unit is month), user age, current terminal brand, current terminal model, current terminal usage start time, brand of previous terminal used before current terminal is used, previous terminal model, previous terminal usage start time, previous terminal usage end time, average usage time of all terminals used by the user, 4G traffic consumption, roaming traffic consumption, total number of calls, number of calling calls, number of called calls, number of local calls, number of roaming calls, number of peer-to-peer short messages, number of idle times, number of idle days, and the like. Of course, it should be noted herein that the data included in the terminal billing data may be some of the above data, and may also be not limited to the above data, that is, the data included in the terminal billing data may be obtained from the charging center and the core network side according to actual needs.
Step 102: and acquiring characteristic data for predicting whether the user has the tendency of changing the machine according to the terminal bill data.
In this step, specifically, after the terminal bill data is obtained, feature data for predicting whether the user has a tendency to change the machine may be obtained according to the terminal bill data, so that whether the user has a tendency to change the machine may be predicted according to the obtained feature data.
Of course, it should be noted herein that the number of the feature data corresponding to each user is multiple, that is, the feature data corresponding to each user is multidimensional data, so as to ensure accuracy when predicting whether the user has a tendency to change a machine according to the feature data.
Step 103: and detecting whether the user has a tendency to change the machine or not according to the characteristic data and a deep neural network prediction model obtained by pre-training.
In this step, specifically, a deep neural network prediction model may be obtained through pre-training, and then whether the user has a tendency to change the machine is detected according to the obtained feature data and the deep neural network prediction model obtained through pre-training, so that whether the user has the tendency to change the machine is predicted by using the deep neural network prediction model, the prediction accuracy is improved, and the success rate of the operator in terminal marketing with the prediction result is improved.
In this way, in the embodiment, by acquiring the terminal bill data of the user in the preset time period, acquiring the feature data for predicting whether the user has the tendency to change the machine according to the terminal bill data, and finally detecting whether the user has the tendency to change the machine according to the acquired feature data and the deep neural network prediction model obtained by pre-training, accurate prediction of whether the user has the tendency to change the machine by using the deep neural network prediction model is realized, the prediction accuracy of the tendency to change the machine of the user is improved, and thus the success rate of the operator in terminal marketing with the prediction result is improved.
Further, in the acquiring of the feature data for predicting whether the subscriber has a tendency to change the telephone, based on the terminal billing data, the feature data may include at least one feature data related to subscriber information, at least one feature data related to terminal information, at least one feature data related to voice and information communication, and at least one feature data related to traffic information. Therefore, the feature data comprises the data of the types, the dimensionality of the feature data is increased, and the prediction accuracy when whether the user has the tendency of changing the machine or not is predicted by the acquired feature data is increased.
In addition, specifically, the at least one feature data related to the user information may include at least one of ARPU, basic monthly rental, online time, and user age; the at least one characteristic data related to the terminal information may include at least one of a used time of a current terminal, a brand of the current terminal, a used time of a previous terminal used by the user before using the current terminal, a brand of the previous terminal, and an average used time of all terminals used by the user; the at least one characteristic data related to the voice and information communication may include at least one of a total number of calls, a total call duration, a number of calling calls, a number of called calls, and a number of point-to-point short messages; the at least one characteristic data related to the traffic information may comprise at least one of total traffic consumption, 4G traffic consumption and roaming traffic consumption.
The used time of the current terminal can be calculated according to the use start time of the current terminal in the terminal bill data, and the used time of the previous terminal can be calculated according to the use start time of the previous terminal and the use end time of the previous terminal in the terminal bill data. In addition, other data in the characteristic data except the used time of the current terminal and the used time of the previous terminal can be directly extracted from the terminal bill data.
In addition, specifically, for quantitative use of the feature data, different terminal brands may be replaced with different numbers to realize quantitative use of the feature data. For example, the first brand corresponding number may be 1, the second brand corresponding number may be 2, the third brand corresponding number may be 3, etc., thereby facilitating the analysis of different characteristic data.
In addition, before detecting whether the user has a tendency to change the machine according to the feature data and the deep neural network prediction model obtained through pre-training, the deep neural network prediction model is further built and obtained through training.
When a deep neural network prediction model is built and trained, the deep neural network model can be built based on tensorflow; the deep neural network model comprises an input layer provided with N neurons, M hidden layers and an output layer, wherein the M hidden layers are respectively provided with a plurality of neurons, and the value of N is the same as the quantity of the characteristic data; training the deep neural network model by adopting training characteristic data corresponding to a first preset number of users with known machine changing tendency, and training to obtain a weight value of the deep neural network model; testing model accuracy of the deep neural network model loaded with the weight value by adopting test characteristic data corresponding to users with known machine changing tendency in a second preset number; and when the model accuracy is greater than a preset threshold value, determining the deep neural network model loaded with the weight value as the deep neural network prediction model.
Specifically, the construction of the deep neural network model based on the tenserflow comprises an input layer provided with N neurons, an output layer and M hidden layers.
Specifically, the value of N in the input layer may be set according to the number of the feature data. For example, when the number of feature data is 19, the value of N is also 19, that is, the input layer includes 19 neurons. In addition, the value of M can be 8, the deep neural network model that builds promptly can include 8 layers, and wherein the first hidden layer that the data was input first and the second hidden layer that data input in proper order all include 128 neurons, and the third hidden layer and fourth hidden layer all include 64 neurons, and the fifth hidden layer and sixth hidden layer all include 32 neurons, and the seventh hidden layer and eighth hidden layer all include 16 neurons. Of course, it should be noted herein that the hidden layers are arranged in the order of data transmission between the layers. In addition, the output layer contains a neuron, namely the data output by the output layer is one-dimensional data used for indicating whether the user has the tendency of changing the machine.
In addition, the activation functions used by the hidden layer are all linear rectification functions (called ReLU for short), and the activation function used by the output layer is sigmoid.
In addition, specifically, when the deep neural network model is trained, desensitized terminal bill data (total 27000 pieces of data) of a plurality of (for example, 9000) known users inclined to change machines in three months can be obtained from the charging center and the core network side, and serve as a data set for deep neural network model training, wherein in the data set, each user is labeled with whether a machine has been changed or not. Then, training feature data corresponding to a first preset number of users and testing feature data corresponding to a second preset number of users can be obtained from the data set, so that the deep neural network model can be trained and tested. For example, the first preset number may be 70% of all known users prone to switching, and the second preset number may be 30% of all known users prone to switching. Of course, it should be noted that the first preset number and the second preset number may be set according to actual requirements, and specific values of the first preset number and the second preset number are not specifically limited herein.
In addition, specifically, the data included in the training feature data and the data included in the testing feature data are both the same as the data included in the feature data corresponding to the user to be tested, that is, when the number of the feature data corresponding to the user to be tested is 19, the training feature data and the testing feature data include the same 19 data, so that the practicability and the testing accuracy of the deep neural network model trained according to the training feature data and tested by the testing feature data are ensured.
In addition, specifically, when the deep neural network model is trained by using training feature data corresponding to a first preset number of users with known tendency to change the machine to obtain a weight value of the deep neural network model, the value of the iteration number (epochs) may be set to 150, and the value of the batch size (batch size) may be set to 10, at this time, the deep neural network model may learn the weight value autonomously by training the training feature data corresponding to the first preset number of users. In addition, in the process of autonomously learning the weight value, the training error is gradually reduced along with the increase of the training times, the model is gradually converged, and after the training for multiple times, the accuracy of the deep neural network model is obviously higher than that of other machine learning algorithms.
At this time, when the weight value of the deep neural network model is obtained through training, the accuracy of the deep neural network model loaded with the weight value can be directly tested through the test characteristic data, and when the model accuracy is greater than a preset threshold value, the deep neural network model loaded with the weight value is determined to be a deep neural network prediction model.
Therefore, the deep neural network prediction model is obtained by training and testing the model accuracy of the deep neural network model, so that whether the user has the tendency to change the machine or not can be predicted through the deep neural network prediction model obtained by training, the accurate prediction of whether the user has the tendency to change the machine or not is realized, and the prediction precision of the tendency to change the machine of the user is improved.
In addition, further, after a deep neural network prediction model is obtained through training, when whether the user has a tendency to change the machine is detected according to the feature data and the deep neural network prediction model obtained through pre-training, the feature data can be directly input into the deep neural network prediction model, and a numerical value output after the deep neural network prediction model analyzes the feature data is obtained; wherein when the value is 1, determining that the user has a tendency to change machines; when the value is 0, determining that the user has no tendency to change the machine. Therefore, whether the user has the tendency of changing the machine or not is accurately predicted through the deep neural network prediction model, and the prediction accuracy of the user for changing the machine is improved.
Specifically, when the numerical value output by the deep neural network prediction model is obtained, the identification code of the user corresponding to the numerical value may be an identification code subjected to desensitization processing, so that privacy protection of the identification code of the user is ensured. Of course, the identification code may be the terminal number of the user.
In addition, further, after detecting whether the user has a tendency to change the machine according to the feature data and a deep neural network prediction model obtained through pre-training, a user list with the tendency to change the machine can be obtained according to a detection result of whether the user has the tendency to change the machine; wherein, the identification code corresponding to the user in the user list is the identification code after desensitization treatment; and calling a capability open platform through an application programming interface (API interface for short) to carry out voice or information communication on the users in the user list, wherein both the voice and the information carry terminal marketing information. Therefore, the voice or information communication is carried out on the users in the user list with the change tendency by calling the capability open platform, and based on the prediction accuracy of the change tendency of the users in the user list, the terminal marketing of the operators to the unnecessary users is avoided, the invalid workload of the operators is reduced, and the success rate of the terminal marketing of the operators is improved.
Specifically, the identification code corresponding to the user in the user list may be an identification code of the user, and may also be a terminal number, that is, the identification code is not specifically limited herein.
In addition, specifically, when the capability opening platform is called through the API interface, the desensitization short message capability and the outbound party capability of the capability opening platform can be called respectively, so as to initiate accurate marketing to the users in the user list.
The desensitization short message capability refers to that a user in a user list sends a short message carrying terminal marketing information. The desensitization short message capability is realized by opening the short message capability of the southbound short message gateway to the northbound application through a capability opening platform. The desensitization short message capability comprises a desensitization short message sending interface, a state report obtaining interface and a state report receiving interface, wherein the desensitization short message sending interface is used for sending a desensitization number to a short message receiving party, the state report obtaining interface is used for obtaining a state report whether the short message is successfully sent to a user from a short message gateway, and the state report receiving interface is used for notifying a short message state report to northbound application. Thus, the number privacy of the user is protected by using the desensitization short message capability.
In addition, the outbound party capability points to the capability opening platform to initiate a one-way voice outbound request, the capability opening platform initiates the outbound party request to the southbound audio and video capability network element, the audio and video capability network element dials the user, and the party initiating the one-way voice outbound request hangs up after the user is connected. The outbound party capability is to initiate a voice outbound request carrying terminal marketing information to the user in the user list. The capability of the calling party is realized by opening the capabilities of a south audio and video capability network element SCP AS and a volte AS to a north application through a capability opening platform, and carrying user numbers in a desensitized user list in a unidirectional voice calling request message. SCP AS is adopted when the called user is 2/3G user, and volteAS is adopted when the called user is volte user.
Of course, it should be noted here that the user list may also be stored on the capability openness platform, so that the third party application can call the capability openness platform through the API interface to perform voice or information communication on the user in the user list.
Therefore, the voice or information communication is carried out on the users in the user list with the change tendency by calling the capacity open platform, the terminal marketing information is carried in the voice and the information, and based on the prediction accuracy of the user with the change tendency in the user list, the terminal marketing of other users without the change tendency is avoided, the unnecessary workload of the operator is reduced, and meanwhile, the success rate of the terminal marketing of the operator is improved.
Therefore, the embodiment of the invention realizes the accurate prediction of whether the user has the tendency of changing the machine by using the deep neural network prediction model, improves the prediction precision of the tendency of changing the machine of the user and further improves the success rate of an operator when carrying out terminal marketing by using the prediction result of the tendency of changing the machine of the user.
As shown in fig. 2, a block diagram of a device for predicting a switch tendency in an embodiment of the present invention is shown, where the device includes:
a first obtaining module 201, configured to obtain terminal bill data of a user within a preset time period;
a second obtaining module 202, configured to obtain, according to the terminal billing data, feature data used for predicting whether the user has a tendency to change machines;
and the detection module 203 is configured to detect whether the user has a tendency to change the machine according to the feature data and a deep neural network prediction model obtained through pre-training.
According to the device for predicting the machine changing tendency provided by the embodiment, the terminal bill data of the user in the preset time period is acquired through the first acquisition module 201, the characteristic data used for predicting whether the user has the machine changing tendency is acquired through the second acquisition module 202 according to the terminal bill data, and finally, whether the user has the machine changing tendency is detected through the detection module 203 according to the characteristic data and the deep neural network prediction model obtained through pre-training, so that the accurate prediction of whether the user has the machine changing tendency by using the deep neural network prediction model is realized, the prediction precision of the machine changing tendency of the user is improved, and the success rate of the operator in terminal marketing with the prediction result of the machine changing tendency of the user is further improved.
Optionally, the feature data comprises: at least one characteristic data related to user information, at least one characteristic data related to terminal information, at least one characteristic data related to voice and information communication, and at least one characteristic data related to traffic information.
Optionally, the apparatus further comprises:
the model training module is used for building and training to obtain a deep neural network prediction model; wherein the content of the first and second substances,
the model training module comprises:
the model building unit is used for building a deep neural network model based on the tenserflow; the deep neural network model comprises an input layer provided with N neurons, M hidden layers and an output layer, wherein the M hidden layers are respectively provided with a plurality of neurons, and the value of N is the same as the quantity of the characteristic data;
the training unit is used for training the deep neural network model by adopting training characteristic data corresponding to a first preset number of users with known machine changing tendency, and training to obtain a weight value of the deep neural network model;
the testing unit is used for carrying out model accuracy testing on the deep neural network model loaded with the weight value by adopting testing characteristic data corresponding to users with known machine changing tendency in a second preset number;
and the determining unit is used for determining the deep neural network model loaded with the weight value as the deep neural network prediction model when the model accuracy is greater than a preset threshold value.
Optionally, the detection module 203 is configured to input the feature data into the deep neural network prediction model, so as to obtain a numerical value output after the deep neural network prediction model analyzes the feature data; wherein the content of the first and second substances,
when the value is 1, determining that the user has a tendency to change machines; when the value is 0, determining that the user has no tendency to change the machine.
Optionally, the apparatus further comprises:
the third acquisition module is used for acquiring a user list with the change tendency according to the detection result of whether the user has the change tendency; wherein, the identification code corresponding to the user in the user list is the identification code after desensitization treatment;
and the communication module is used for calling the capability open platform through an Application Programming Interface (API) interface and carrying out voice or information communication on the users in the user list, wherein both the voice and the information carry terminal marketing information.
Therefore, the device for predicting the machine changing tendency provided by the embodiment of the invention detects whether the user has the machine changing tendency by acquiring the terminal bill data of the user in the preset time period, acquiring the characteristic data for predicting whether the user has the machine changing tendency according to the terminal bill data, and finally detecting whether the user has the machine changing tendency according to the acquired characteristic data and the deep neural network prediction model obtained by pre-training, so that the accurate prediction of whether the user has the machine changing tendency by using the deep neural network prediction model is realized, the prediction accuracy of the machine changing tendency of the user is improved, and the success rate of the operator in terminal marketing according to the prediction result of the machine changing tendency of the user is further improved.
It should be noted that, in the embodiment of the present invention, the related functional modules may be implemented by a hardware processor (hardware processor), and the same technical effect can be achieved, which is not described herein again.
In yet another embodiment of the present invention, an electronic device is provided, as shown in fig. 3, which includes a memory (memory)301, a processor (processor)302, and a computer program stored on the memory 301 and executable on the processor 302. The memory 301 and the processor 302 complete communication with each other through the bus 303. The processor 302 is configured to call program instructions in the memory 301 to perform the following method: acquiring terminal bill data of a user in a preset time period; acquiring characteristic data for predicting whether the user has the tendency of changing the machine or not according to the terminal bill data; and detecting whether the user has a tendency to change the machine or not according to the feature data and a deep neural network prediction model obtained by pre-training.
The electronic device provided by the embodiment of the invention can execute specific steps in the change tendency prediction method and can achieve the same technical effect, and the specific steps are not described herein.
Further, the program instructions in the memory 301 may be implemented in the form of software functional units and may be stored in a computer-readable storage medium when sold or used as a stand-alone product. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
In a further embodiment of the invention, a non-transitory computer readable storage medium is provided, having stored thereon a computer program which, when executed by a processor, is operative to perform the method of: acquiring terminal bill data of a user in a preset time period; acquiring characteristic data for predicting whether the user has the tendency of changing the machine or not according to the terminal bill data; and detecting whether the user has a tendency to change the machine or not according to the feature data and a deep neural network prediction model obtained by pre-training.
The non-transitory computer-readable storage medium provided by the embodiment of the invention can execute specific steps in the change tendency prediction method and can achieve the same technical effect, and the specific steps are not described herein.
In yet another embodiment of the present invention, a computer program product is provided, the computer program product comprising a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions that when executed by a computer perform the method of: acquiring terminal bill data of a user in a preset time period; acquiring characteristic data for predicting whether the user has the tendency of changing the machine or not according to the terminal bill data; and detecting whether the user has a tendency to change the machine or not according to the feature data and a deep neural network prediction model obtained by pre-training.
The computer program product provided by the embodiment of the invention can execute specific steps in the method for predicting the tendency of the machine change, and can achieve the same technical effect, and the specific description is not provided herein.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and 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.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (8)
1. A method for predicting a tendency to change a machine, the method comprising:
acquiring terminal bill data of a user in a preset time period;
acquiring characteristic data for predicting whether the user has the tendency of changing the machine or not according to the terminal bill data;
and detecting whether the user has a tendency to change the machine or not according to the feature data and a deep neural network prediction model obtained by pre-training.
2. The method of claim 1, wherein the characterization data comprises: at least one characteristic data related to user information, at least one characteristic data related to terminal information, at least one characteristic data related to voice and information communication, and at least one characteristic data related to traffic information.
3. The method according to claim 1, wherein before detecting whether the user has a tendency to change the machine based on the feature data and a pre-trained deep neural network prediction model, the method further comprises:
building and training to obtain a deep neural network prediction model; wherein the content of the first and second substances,
the building and training to obtain the deep neural network prediction model comprises the following steps:
building a deep neural network model based on the tensoflow; the deep neural network model comprises an input layer provided with N neurons, M hidden layers and an output layer, wherein the M hidden layers are respectively provided with a plurality of neurons, and the value of N is the same as the quantity of the characteristic data;
training the deep neural network model by adopting training characteristic data corresponding to a first preset number of users with known machine changing tendency, and training to obtain a weight value of the deep neural network model;
testing model accuracy of the deep neural network model loaded with the weight value by adopting test characteristic data corresponding to users with known machine changing tendency in a second preset number;
and when the model accuracy is greater than a preset threshold value, determining the deep neural network model loaded with the weight value as the deep neural network prediction model.
4. The method according to claim 1, wherein the detecting whether the user has a tendency to change a machine according to the feature data and a deep neural network prediction model trained in advance comprises:
inputting the characteristic data into the deep neural network prediction model to obtain a numerical value output after the deep neural network prediction model analyzes the characteristic data; wherein the content of the first and second substances,
when the value is 1, determining that the user has a tendency to change machines; when the value is 0, determining that the user has no tendency to change the machine.
5. The method according to claim 1, wherein after detecting whether the user has a tendency to change the machine based on the feature data and a pre-trained deep neural network prediction model, the method further comprises:
acquiring a user list with the change tendency according to the detection result of whether the user has the change tendency; wherein, the identification code corresponding to the user in the user list is the identification code after desensitization treatment;
and calling the capability open platform through an Application Programming Interface (API) interface to carry out voice or information communication on the users in the user list, wherein both the voice and the information carry terminal marketing information.
6. A device for predicting a tendency to change a machine, the device comprising:
the first acquisition module is used for acquiring terminal bill data of a user in a preset time period;
the second acquisition module is used for acquiring characteristic data for predicting whether the user has the tendency of changing the machine or not according to the terminal bill data;
and the detection module is used for detecting whether the user has a tendency to change the machine or not according to the feature data and a deep neural network prediction model obtained by pre-training.
7. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the method for predicting the tendency of a machine change as claimed in any one of claims 1 to 5 when executing the computer program.
8. A non-transitory computer readable storage medium, having stored thereon a computer program, wherein the computer program, when executed by a processor, implements the steps of the method for predicting change tendency according to any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810595922.XA CN110648153A (en) | 2018-06-11 | 2018-06-11 | Change tendency prediction method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810595922.XA CN110648153A (en) | 2018-06-11 | 2018-06-11 | Change tendency prediction method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110648153A true CN110648153A (en) | 2020-01-03 |
Family
ID=68988341
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810595922.XA Pending CN110648153A (en) | 2018-06-11 | 2018-06-11 | Change tendency prediction method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110648153A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104794195A (en) * | 2015-04-17 | 2015-07-22 | 南京大学 | Data mining method for finding potential telecommunication users changing cell phones |
| CN106845731A (en) * | 2017-02-20 | 2017-06-13 | 重庆邮电大学 | A kind of potential renewal user based on multi-model fusion has found method |
| CN106991577A (en) * | 2016-01-21 | 2017-07-28 | 中国移动通信集团湖南有限公司 | A kind of method and device for determining targeted customer |
| CN107016460A (en) * | 2017-03-27 | 2017-08-04 | 中国联合网络通信集团有限公司广西壮族自治区分公司 | User changes planes Forecasting Methodology and device |
| CN107808132A (en) * | 2017-10-23 | 2018-03-16 | 重庆邮电大学 | A kind of scene image classification method for merging topic model |
-
2018
- 2018-06-11 CN CN201810595922.XA patent/CN110648153A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104794195A (en) * | 2015-04-17 | 2015-07-22 | 南京大学 | Data mining method for finding potential telecommunication users changing cell phones |
| CN106991577A (en) * | 2016-01-21 | 2017-07-28 | 中国移动通信集团湖南有限公司 | A kind of method and device for determining targeted customer |
| CN106845731A (en) * | 2017-02-20 | 2017-06-13 | 重庆邮电大学 | A kind of potential renewal user based on multi-model fusion has found method |
| CN107016460A (en) * | 2017-03-27 | 2017-08-04 | 中国联合网络通信集团有限公司广西壮族自治区分公司 | User changes planes Forecasting Methodology and device |
| CN107808132A (en) * | 2017-10-23 | 2018-03-16 | 重庆邮电大学 | A kind of scene image classification method for merging topic model |
Non-Patent Citations (1)
| Title |
|---|
| 杨秀松: "基于数据挖掘的电信潜在换机客户的预测研究", 《中国优秀博硕士学位论文全文数据库(硕士)信息科技辑》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107566358B (en) | Risk early warning prompting method, device, medium and equipment | |
| CN109429230B (en) | Communication fraud identification method and system | |
| CN109302719B (en) | LTE cell capacity prediction analysis method and device | |
| CN106294508B (en) | Brushing amount tool detection method and device | |
| US9357362B2 (en) | System and method for fast and accurate detection of SMS spam numbers via monitoring grey phone space | |
| CN109446171B (en) | Data processing method and device | |
| WO2020134523A1 (en) | User identification method and device | |
| CN108764951B (en) | User similarity obtaining method and device, equipment and storage medium | |
| CN109168168B (en) | Method for detecting international embezzlement | |
| CN110019382B (en) | User intimacy index determination method and device, storage medium and electronic equipment | |
| CN112385267B (en) | Method and device for determining target cell of UE, communication equipment and storage medium | |
| CN115470867A (en) | Agent matching method, device, equipment and storage medium based on knowledge graph | |
| CN105162931B (en) | The sorting technique and device of a kind of communicating number | |
| CN110648153A (en) | Change tendency prediction method and device, electronic equipment and storage medium | |
| CN110516151B (en) | Effective behavior detection and personalized recommendation method | |
| CN111368858B (en) | User satisfaction evaluation method and device | |
| CN107306419A (en) | A kind of end-to-end quality appraisal procedure and device | |
| CN109951609B (en) | Malicious telephone number processing method and device | |
| CN116800886A (en) | Abnormal number identification method and device, storage medium and electronic equipment | |
| EP4093003A1 (en) | Spoofed telephone call identifier | |
| CN112307075B (en) | User relationship identification method and device | |
| CN110856159B (en) | Method, device and storage medium for determining family circle members | |
| US20220172077A1 (en) | Systems and methods for automated and dynamic journey intervention using machine learning techniques | |
| CN108235268B (en) | Method and device for acquiring use times of IMEI (international mobile equipment identity) of call terminal | |
| CN108509560B (en) | User similarity obtaining method and device, equipment and storage medium |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200103 |
|
| RJ01 | Rejection of invention patent application after publication |