User data analysis and intelligent network selection combined method and system

Abstract

The invention relates to a user data analysis and intelligent network selection combined method and system, and belongs to the technical field of data processing, and the method comprises the steps: employing a Markov chain to predict the movement track information of a user in a preset time, and updating the residence prediction value of the user residing in a cell list; determining a target user according to the residence prediction value of the user, constructing a clustering model for data of a normal type of behavior information of the target user by adopting a K-means + + algorithm, and generating a normal information matrix; determining an abnormal information matrix, and identifying real abnormal information according to the abnormal information matrix and the normal information matrix; and training a user input model in the real abnormal information to obtain information of each user reporting base station network element and real abnormal information data. According to the method, the real abnormal information data reported by the user to the network element of the base station is accurately obtained.

Description

Technical Field

[0001] The present invention belongs to the technical field of data processing, and in particular relates to a method and system for combining user data analysis with intelligent network selection. Background Art

[0002] In the related technologies, in mobile communication networks, with the continuous increase in the number of users and the enhancement of mobility, the management and optimization of network resources have become increasingly complex. In order to effectively plan the network, improve resource utilization, ensure service quality, and promptly discover and handle potential network failures or abnormal user behavior, operators need to accurately predict and analyze user movement trajectories and behavior information. Summary of the invention

[0003] In view of the above deficiencies in the prior art, an object of the present invention is to provide a method and system for combining user data analysis with intelligent network selection.

[0004] The first aspect of the present invention proposes a method for combining user data analysis with intelligent network selection, including: S1, using Markov chain to predict the user's movement trajectory information within a preset time, and updating the residence prediction value of the user residing in the cell list; S2, determining the target user according to the user's residence prediction value, and using the K-means++ algorithm to build a clustering model for the normal type of data of the target user's behavior information to generate a normal information matrix; S3, determining the abnormal information matrix, and identifying the real abnormal information according to the abnormal information matrix and the normal information matrix; S4, training the user input model in the real abnormal information to obtain the information reported by each user to the base station network element and the real abnormal information data.

[0005] Further, determining the target user according to the residence prediction value of the user includes: determining a target residence prediction value greater than a preset threshold among the residence prediction values ​​of the user; and taking the user corresponding to the target residence prediction value as the target user.

[0006] Furthermore, the data of the normal type of the target user behavior information includes multiple feature data, wherein the K-means++ algorithm is used to construct a clustering model for the data of the normal type of the target user behavior information to generate a normal information matrix, including: randomly selecting one feature data as the center point from the multiple feature data; calculating the distance from the point of each feature data to the center point; selecting a new center point according to the distance and weighted probability distribution until N center points are selected; and using standard K-means clustering to classify the multiple feature data to generate the normal information matrix.

[0007] Furthermore, the cluster model formula includes: ||Xθ-y|| 2 +||Γθ|| 2,θ(a)=(X T X+aI) - 1 X T y, where X represents the input, y represents the predicted output, || represents the regularization operation, I represents the identity matrix, θ represents the fitting hyperparameter, Γ represents the weight constant, a represents the weight of the identity matrix, θ(a) represents the value of θ when a is determined, and T represents the transpose.

[0008] Furthermore, according to the abnormal information matrix and the normal information matrix, real abnormal information is identified, including: decomposing the user-classification interaction matrix R into two low-rank matrices U and V, and approximating R by the product of the matrices U and V; adopting the idea of ​​linear regression to minimize the square error between the original score and the predicted score to obtain the objective function; and comparing the similarity between the abnormal information matrix and the normal information matrix when merging the abnormal information matrix and the normal information matrix to identify real abnormal information.

[0009] Further, according to The objective function is obtained, where U i represents the user latent vector, V j represents the classification latent vector, represents the matrix norm, λ represents the regularization factor, K represents the (i, j) pair with known scores, T represents the transpose, U and V represent low-rank matrices, and R represents the product of matrices.

[0010] Furthermore, it also includes: outputting the information reported by each user to the base station network element and the real abnormal information data to the network side.

[0011] The second aspect of the present invention proposes a system combining user data analysis with intelligent network selection, including: an update module, which uses a Markov chain to predict the user's movement trajectory information within a preset time, and updates the residence prediction value of the user residing in the cell list; a generation module, which is used to determine the target user according to the user's residence prediction value, and use the K-means++ algorithm to build a clustering model for the normal type of data of the target user's behavior information to generate a normal information matrix; an identification module, which is used to determine the abnormal information matrix, and identify the real abnormal information based on the abnormal information matrix and the normal information matrix; an acquisition module, which is used to train the user input model in the real abnormal information, and obtain the information reported by each user to the base station network element and the real abnormal information data.

[0012] According to a third aspect of the present invention, an electronic device is proposed, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute any of the methods described in the first aspect of the present invention.

[0013] According to a fourth aspect of the present invention, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable the computer to execute any one of the methods according to the first aspect of the present invention.

[0014] The beneficial effects of the present invention are as follows:

[0015] The method and system for combining user data analysis with intelligent network selection described in the present invention use Markov chains to predict the user's movement trajectory information within a preset time, and update the residence prediction value of the user residing in the cell list; determine the target user according to the user's residence prediction value, and use the K-means++ algorithm to build a clustering model for the normal type of target user behavior information data to generate a normal information matrix; determine the abnormal information matrix, and identify the real abnormal information based on the abnormal information matrix and the normal information matrix; train the user input model in the real abnormal information to obtain the information reported by each user to the base station network element and the real abnormal information data. This method predicts the user's movement trajectory, constructs a normal behavior information matrix and compares it with the abnormal information matrix to identify the real abnormal behavior, and performs model training on the user data in the real abnormal information, thereby accurately obtaining the real abnormal information data reported by the user to the base station network element. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings are only used to illustrate specific embodiments and are not considered to limit the present invention. In the entire drawings, the same reference symbols represent the same components. Obviously, the drawings described below are only some embodiments recorded in the embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0017] Figure 1 is a flow chart of a method for combining user data analysis with intelligent network selection according to an embodiment of the present invention;

[0018] Figure 2 is a flow chart of a method for combining user data analysis with intelligent network selection according to a specific embodiment of the present invention;

[0019] Figure 3 is a schematic diagram of a system combining user data analysis with intelligent network selection according to an embodiment of the present invention;

[0020] Figure 4It is a structural block diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0021] In order to enable those skilled in the art to better understand the technical solutions in the embodiments of the present invention, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. It should be understood that these descriptions are only exemplary and are not intended to limit the scope of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in the field without making creative work should fall within the scope of protection of the present invention.

[0022] Furthermore, in the following description, descriptions of well-known structures and technologies are omitted to avoid unnecessarily obscuring the concepts disclosed in the present invention.

[0023] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the orientation or position relationship indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inside", "outside", etc. is based on the orientation or position relationship shown in the drawings, which is only for the convenience of describing the present invention and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation of the present invention. In addition, the terms "first", "second", and "third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. The terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be a connection between the two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0024] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Instead, they are merely examples of methods and systems consistent with some aspects of the present invention as detailed in the appended claims.

[0025] The present invention proposes a method, system and related equipment for combining user data analysis with intelligent network selection. Specifically, the method, system and related equipment for combining user data analysis with intelligent network selection according to an embodiment of the present invention are described below with reference to the accompanying drawings.

[0026] Figure 1This is a flow chart of a method for combining user data analysis with intelligent network selection according to an embodiment of the present invention. It should be noted that the method for combining user data analysis with intelligent network selection in the embodiment of the present invention can be applied to the system for combining user data analysis with intelligent network selection in the embodiment of the present invention. The system for combining user data analysis with intelligent network selection can be configured on an electronic device or in a server. This embodiment of the present application does not limit this.

[0027] like Figure 1 As shown, the method of combining user data analysis with intelligent network selection includes:

[0028] S110, using a Markov chain to predict the movement trajectory information of the user within a preset time, and updating the residence prediction value of the user who resides in the cell list.

[0029] In the embodiments of the present invention, the user's historical movement data can be used in combination with the state transition characteristics of the Markov chain to predict the user's future movement trajectory, providing a basis for subsequent behavior pattern analysis and anomaly detection. The specific implementation method can refer to the subsequent embodiments.

[0030] S120, determining the target user according to the user's predicted residence value, building a clustering model for the target user's behavior information normal type data using the K-means++ algorithm, and generating a normal information matrix.

[0031] In the embodiment of the present invention, a target residence prediction value greater than a preset threshold value is determined among the residence prediction values ​​of the users; the user corresponding to the target residence prediction value is taken as the target user, and then a clustering model is constructed using the K-means++ algorithm for the data of the normal type of the target user's behavior information to generate a normal information matrix. The specific implementation method can be referred to the subsequent embodiments.

[0032] S130, determining an abnormal information matrix, and identifying real abnormal information according to the abnormal information matrix and the normal information matrix.

[0033] In an embodiment of the present invention, the abnormal behavior information of users accessed from the network side is used as an abnormal information matrix and merged with the normal information matrix. Through the matrix decomposition method, this information can be merged based on the similarity of the normal information matrix, so that the false abnormal information that may be generated by false alarms can be aggregated into the normal information matrix. Among them, the abnormal information with low similarity is regarded as the real abnormal information. The specific implementation method can refer to the above embodiment.

[0034] S140, training a user input model in the real abnormal information to obtain information reported by each user to the base station network element and real abnormal information data.

[0035] In an embodiment of the present invention, when the information reported by each user on the base station network element and the real abnormal information data are obtained, the information reported by each user on the base station network element and the real abnormal information data are output to the network side.

[0036] According to the method of combining user data analysis with intelligent network selection in an embodiment of the present invention, a Markov chain is used to predict the user's movement trajectory information within a preset time, and the residence prediction value of the user residing in the cell list is updated; the target user is determined according to the user's residence prediction value, and the K-means++ algorithm is used to build a clustering model for the normal type of target user behavior information data to generate a normal information matrix; the abnormal information matrix is ​​determined, and the real abnormal information is identified based on the abnormal information matrix and the normal information matrix; the user input model in the real abnormal information is trained to obtain the information reported by each user to the base station network element and the real abnormal information data. This method predicts the user's movement trajectory, constructs a normal behavior information matrix and compares it with the abnormal information matrix to identify the real abnormal behavior, and performs model training on the user data in the real abnormal information, thereby accurately obtaining the real abnormal information data reported by the user to the base station network element.

[0037] In order to make it easier for those skilled in the art to understand the present invention, Figure 2 A method for combining user data analysis with intelligent network selection according to a specific embodiment of the present invention is as follows. Figure 2 As shown, the method of combining user data analysis with intelligent network selection includes:

[0038] S210, using a Markov chain to predict the movement trajectory information of the user within a preset time, and updating the residence prediction value of the user who resides in the cell list.

[0039] In an embodiment of the present invention, historical movement trajectory data of users is collected, including the residence time of users at different times and different places (such as cells). Construct a state transition matrix: Based on historical data, construct a state transition matrix of users between different cells. Utilize the predictive properties of the Markov chain, predict the movement trajectory and residence probability of users in the future period of time based on the current state (the cell where the user is currently located) and the state transition matrix. Based on the prediction results, update the residence prediction value of users residing in the cell list.

[0040] S220: Determine a target user according to the user's predicted residence value.

[0041] In an embodiment of the present invention, a target residence prediction value greater than a preset threshold value among the residence prediction values ​​of the users is determined; and a user corresponding to the target residence prediction value is designated as a target user.

[0042] S230, using the K-means++ algorithm to construct a clustering model for the normal type of target user behavior information data to generate a normal information matrix.

[0043] Among them, the normal type of data of target user behavior information includes a variety of characteristic data, including but not limited to basic base station data (such as antenna height, longitude and latitude, azimuth, etc.; cell frequency, neighboring cell relationship, etc.), base station networking data (such as single carrier, multi-carrier, coverage layer, capacity layer, etc.), actual coverage scene information (high-rise buildings, high-speed railways, subways), performance indicators such as traffic, load and switching information, measurement report data, UE scene mall, subway, community, highway and residence, etc., user level information served by each community, etc.

[0044] In an embodiment of the present invention, one type of feature data is randomly selected from a plurality of feature data as a center point; the distance from a point of each type of feature data to the center point is calculated; a new center point is selected based on the distance and weighted probability distribution until N center points are selected; and the plurality of feature data are classified using standard K-means clustering to generate a normal information matrix.

[0045] In order to ensure the stability and accuracy of the clustering model, by adding a regularization term, the model can be prevented from overfitting the training data during the training process, thereby improving the generalization ability of the model. That is, the clustering model fitting formula includes: The clustering model fitting formula includes: ||Xθ-y|| 2 +||Γθ|| 2 ,θ(a)=(X T X+aI) -1 X T y, where X represents the input, y represents the predicted output, || represents the regularization operation, I represents the identity matrix, θ represents the fitting hyperparameter, Γ represents the weight constant, a represents the weight of the identity matrix, θ(a) represents the value of θ when a is determined, and T represents the transpose.

[0046] S240, determining an abnormal information matrix.

[0047] In an embodiment of the present invention, the abnormal behavior information of users accessed by the network side is used as an abnormal information matrix.

[0048] S250, identifying real abnormal information according to the abnormal information matrix and the normal information matrix.

[0049] In an embodiment of the present invention, the user-classification interaction matrix R is decomposed into two low-rank matrices U and V, and R is approximated by the product of matrices U and V; the idea of ​​linear regression is adopted to minimize the square error between the original score and the predicted score to obtain the objective function; when the abnormal information matrix and the normal information matrix are combined, the similarity between the abnormal information matrix and the normal information matrix is ​​compared to identify the real abnormal information.

[0050] Among them, according to Get the objective function, where U i represents the user latent vector, V j represents the classification latent vector, represents the matrix norm, λ represents the regularization factor, K represents the (i, j) pair with known scores, T represents the transpose, U and V represent low-rank matrices, and R represents the product of matrices.

[0051] For example, suppose there are m users, n categories, and an m*n user-category interaction matrix R, where the values ​​in R represent specific rating values. The idea of ​​matrix decomposition is that R can be decomposed into two low-rank matrices U and V, and then the product of the two is used to approximate R, which is formulated as follows: R≈U T V, where U∈R K×m , V∈R K×n , and k = min(m, n). Column vector U i and V j Represent the user latent vector and classification latent vector respectively.

[0052] To learn the latent vector, the idea of ​​linear regression is used to minimize the square error between the original score and the predicted score, and the objective function is obtained, that is,

[0053] It should be noted that in the matrix decomposition method, the objective function is used to minimize the square error between the original score and the predicted score, while adding a regularization term to prevent overfitting. The optimization process of the objective function (such as using the gradient descent method or the ALS method) aims to find the latent vectors U and V that minimize the error. When merging the normal and abnormal information matrices, the similarity judgment is used to distinguish between false abnormal information and real abnormal information. By comparing the similarity between the abnormal information matrix and the normal information matrix (based on the latent vector obtained by matrix decomposition), abnormal information that is significantly different from the normal behavior pattern, that is, real abnormal information, can be identified. Similarity judgment depends on the results of matrix decomposition, that is, the latent vectors U and V, which reflect the potential relationship between users and classifications. That is, the optimization process of the objective function (matrix decomposition) provides a basis for similarity judgment, so that false abnormal information and real abnormal information can be accurately distinguished. By minimizing the error and preventing overfitting, the objective function ensures the accuracy of the matrix decomposition, thereby improving the reliability of similarity judgment.

[0054] S260, training a user input model in the real abnormal information to obtain information reported by each user to the base station network element and real abnormal information data.

[0055] S270, output the information reported by each user to the base station network element and the real abnormal information data to the network side.

[0056] In the embodiment of the present invention, the users in the real abnormal information are put into the model as dynamic parameters for training, and the information reported by each UE to the base station network element and the real abnormal information data are obtained. These data will be output to the network side for subsequent network optimization and resource allocation.

[0057] According to the method of combining user data analysis with intelligent network selection in an embodiment of the present invention, by collecting the historical movement trajectory data of the user and predicting using the Markov chain, the method can accurately predict the movement trajectory and residence probability of the user within a preset time. This helps the network side to understand the possible location of the user in advance and provide data support for intelligent network selection and network optimization; according to the user's residence prediction value, the method can screen out the target users, that is, those who may frequently reside in a specific cell in the future. This helps the network side to provide more detailed services and resource allocation to key users; the K-means++ algorithm is used to cluster the normal type of data of the target user's behavior information to generate a normal information matrix. By adding regularization terms, the method ensures the stability and accuracy of the clustering model and avoids overfitting of the model during the training process; combined with the abnormal information matrix and the normal information matrix, the method can identify real abnormal information. Through matrix decomposition and similarity judgment, the method can distinguish between false abnormal information and real abnormal information, and provide accurate abnormal information data for the network side; the user in the real abnormal information is put into the model as a dynamic parameter for training, and the information reported by each user to the base station network element and the real abnormal information data are obtained. These data will be output to the network side in real time for subsequent network optimization and resource allocation, improving the intelligence and dynamic level of the network; this method not only provides user movement trajectory prediction and abnormal information identification, but also provides comprehensive network optimization support for the network side by outputting the information reported by each user to the base station network element and the real abnormal information data. This helps the network side better understand user needs and network conditions, thereby performing more accurate network optimization and resource allocation.

[0058] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods of each embodiment of the present application.

[0059] According to one aspect of an embodiment of the present invention, a system combining user data analysis with intelligent network selection is also proposed. Figure 3 is a schematic diagram of a system combining user data analysis with intelligent network selection according to an embodiment of the present invention; Figure 3 As shown, including:

[0060] An updating module 310 is used to predict the movement trajectory information of the user within a preset time by using a Markov chain, and update the residence prediction value of the user residing in the cell list;

[0061] A generation module 320 is used to determine a target user according to the predicted residence value of the user, and to construct a clustering model using a K-means++ algorithm for the normal type of the target user behavior information data to generate a normal information matrix;

[0062] An identification module 330 is used to determine an abnormal information matrix, and identify real abnormal information according to the abnormal information matrix and the normal information matrix;

[0063] The acquisition module 340 is used to train the user input model in the real abnormal information to obtain the information reported by each user to the base station network element and the real abnormal information data.

[0064] According to the user data analysis combined with the intelligent network selection system of the embodiment of the present invention, a Markov chain is used to predict the user's movement trajectory information within a preset time, and the residence prediction value of the user residing in the cell list is updated; the target user is determined according to the user's residence prediction value, and the K-means++ algorithm is used to build a clustering model for the normal type of target user behavior information data to generate a normal information matrix; the abnormal information matrix is ​​determined, and the real abnormal information is identified based on the abnormal information matrix and the normal information matrix; the user input model in the real abnormal information is trained to obtain the information reported by each user to the base station network element and the real abnormal information data. Therefore, by predicting the user's movement trajectory, building a normal behavior information matrix and comparing it with the abnormal information matrix to identify the real abnormal behavior, and training the model for the user data in the real abnormal information, the real abnormal information data reported by the user to the base station network element can be accurately obtained.

[0065] Optionally, the generating module 320 is specifically configured to determine a target residence prediction value greater than a preset threshold among the residence prediction values ​​of the users; and use the user corresponding to the target residence prediction value as the target user.

[0066] Optionally, the normal type of target user behavior information data includes multiple feature data, wherein the generation module 320 is specifically used to randomly select one feature data from the multiple feature data as the center point; calculate the distance from each feature data point to the center point; select a new center point according to the distance and weighted probability distribution until N center points are selected; and use standard K-means clustering to classify the multiple feature data to generate the normal information matrix.

[0067] Optionally, fitting the cluster model formula includes: ||Xθ-y|| 2 +||Γθ|| 2 ,θ(a)=(X T X+aI) -1 X T y, where X represents the input, y represents the predicted output, || represents the regularization operation, I represents the identity matrix, θ represents the fitting hyperparameter, Γ represents the weight constant, a represents the weight of the identity matrix, θ(a) represents the value of θ when a is determined, and T represents the transpose.

[0068] Optionally, the identification module 330 is specifically used to decompose the user-classification interaction matrix R into two low-rank matrices U and V, and approximate R by the product of the matrices U and V; adopt the idea of ​​linear regression to minimize the square error between the original score and the predicted score to obtain the objective function; and when merging the abnormal information matrix and the normal information matrix, compare the similarity between the abnormal information matrix and the normal information matrix to identify the real abnormal information.

[0069] Optionally, the identification module 330 is specifically configured to: The objective function is obtained, where U i represents the user latent vector, V j represents the classification latent vector, represents the matrix norm, λ represents the regularization factor, K represents the (i, j) pair with known scores, T represents the transpose, U and V represent low-rank matrices, and R represents the product of matrices.

[0070] Optionally, an output module is further included, which is used to output the information reported by each user to the base station network element and the real abnormal information data to the network side. According to one aspect of an embodiment of the present invention, an electronic device is provided.

[0071] According to one aspect of an embodiment of the present invention, an electronic device is provided.

[0072] Figure 4 Schematic diagram of the structure of an electronic device according to an embodiment of the present invention. Figure 4 As shown, the electronic device may include one or more ( Figure 4Only one is shown in the figure) processor 102 (processor 102 may include but is not limited to a microprocessor (Microprocessor Unit, referred to as MPU) or a programmable logic device (Programmable logic device, referred to as PLD)) and a memory 104 for storing data. In an exemplary embodiment, the electronic device may also include a transmission device 106 and an input / output device 108 for communication functions. It can be understood by those skilled in the art that Figure 4 The structure shown is for illustration only and does not limit the structure of the above terminal device. Figure 4 More or fewer components as shown, or with Figure 4 Equivalent functions or comparisons shown Figure 4 A different configuration with more features is shown.

[0073] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the method of combining user data analysis with intelligent network selection in the embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, to implement the above method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the terminal device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0074] The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the switching device. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0075] The present invention provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable the computer to execute a method combining user data analysis with intelligent network selection.

[0076] The applicant of the present invention has made a detailed explanation and description of the implementation examples of the present invention in conjunction with the drawings in the specification. However, those skilled in the art should understand that the above implementation examples are only preferred implementation schemes of the present invention, and the detailed description is only to help readers better understand the spirit of the present invention, but not to limit the scope of protection of the present invention. On the contrary, any improvements or modifications based on the inventive spirit of the present invention should fall within the scope of protection of the present invention.

[0077] In addition, each functional unit in each embodiment of the present invention may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0078] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present invention. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present invention.

[0079] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the embodiments of the present invention, rather than to limit them. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the above embodiments, or replace some of the technical features therein by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention. Any changes or replacements that can be easily thought of by those skilled in the art within the technical scope disclosed in the present invention should be covered within the protection scope of the present invention.

Claims

1. A method for combining user data analysis with intelligent network selection, characterized in that: include: S1, using Markov chain to predict the movement trajectory information of users within a preset time, and updating the residence prediction value of users residing in the cell list; S2, determining the target user according to the predicted residence value of the user, building a clustering model for the normal type of behavior information of the target user using the K-means++ algorithm, and generating a normal information matrix; S3, determining an abnormal information matrix, and identifying real abnormal information according to the abnormal information matrix and the normal information matrix; S4, training the user input model in the real abnormal information to obtain the information reported by each user to the base station network element and the real abnormal information data.

2. The method of combining user data analysis with intelligent network selection according to claim 1, characterized in that: Determining a target user according to the predicted residence value of the user includes: Determine a target residence prediction value that is greater than a preset threshold among the residence prediction values ​​of the user; The user corresponding to the target residence prediction value is used as the target user.

3. The method of combining user data analysis with intelligent network selection according to claim 1, characterized in that: The normal type of target user behavior information data includes a variety of feature data, wherein the K-means++ algorithm is used to construct a clustering model for the normal type of target user behavior information data to generate a normal information matrix, including: Randomly selecting one type of feature data from the multiple types of feature data as a center point; Calculate the distance from each feature data point to the center point; Select a new center point based on the distance and weighted probability distribution until N center points are selected; The various feature data are classified using standard K-means clustering to generate the normal information matrix.

4. The method of combining user data analysis with intelligent network selection according to claim 1, characterized in that: The cluster model fitting formula includes: ||Xθ-y|| 2 +||Γθ|| 2 ,θ(a)=(X T X+aI) -1 X T y, where X represents the input, y represents the predicted output, || represents the regularization operation, I represents the identity matrix, θ represents the fitting hyperparameter, Γ represents the weight constant, a represents the weight of the identity matrix, θ(a) represents the value of θ when a is determined, and T represents the transpose.

5. The method of combining user data analysis with intelligent network selection according to claim 1, characterized in that: According to the abnormal information matrix and the normal information matrix, identifying real abnormal information includes: Decompose the user-class interaction matrix R into two low-rank matrices U and V, and approximate R by the product of the matrices U and V; Adopting the idea of ​​linear regression, we minimize the square error between the original score and the predicted score to get the objective function; When the abnormal information matrix and the normal information matrix are combined, the similarity between the abnormal information matrix and the normal information matrix is ​​compared to identify the real abnormal information.

6. The method of combining user data analysis with intelligent network selection according to claim 1, characterized in that: according to The objective function is obtained, where U i represents the user latent vector, V j represents the classification latent vector, represents the matrix norm, λ represents the regularization factor, K represents the (i, j) pair with known scores, T represents the transpose, U and V represent low-rank matrices, and R represents the product of matrices.

7. The method of combining user data analysis with intelligent network selection according to claim 1, characterized in that: Also includes: The information reported by each user to the base station network element and the real abnormal information data are output to the network side.

8. A system combining user data analysis with intelligent network selection, characterized in that: include: An updating module, used to predict the movement trajectory information of users within a preset time by using a Markov chain, and update the residence prediction value of users residing in the cell list; A generation module, used to determine the target user according to the predicted residence value of the user, build a clustering model for the normal type of the target user behavior information data using the K-means++ algorithm, and generate a normal information matrix; An identification module, used to determine an abnormal information matrix, and identify real abnormal information according to the abnormal information matrix and the normal information matrix; The acquisition module is used to train the user input model in the real abnormal information to obtain the information reported by each user to the base station network element and the real abnormal information data.

9. An electronic device, characterized in that: include: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to cause the computer to execute the method according to any one of claims 1 to 7.

Images