An improved method and system for pattern recognition network evaluation

Abstract

The present invention provides an improved method and system for pattern recognition network evaluation, which establishes a network security parameter set, combines historical data items to form a training sample set, extracts two samples in sequence, constructs sample pairs, uses a branched convolutional neural network to eliminate sample pairs that do not belong to the same category, obtains a second training sample set, calls a pattern recognition method to train the second sample set, and then solves the optimal solution to obtain the required vector width, and finally establishes an evaluation model, thereby reducing the impact of the number of samples.

Description

Technical Field

[0001] The present application relates to the field of network multimedia, and in particular to an improved method and system for pattern recognition network evaluation. Background Art

[0002] Existing network security assessments mostly use cluster analysis methods to build models. However, in real scenarios, parameters are not fixedly corresponding to the network environment, resulting in low practical application value of existing assessment methods. If we only improve the corresponding relationship between scenario parameters and network environment, it will still be affected by the number of samples, thus affecting the accuracy of the assessment. Repeated assessments will also bring a lot of computational complexity.

[0003] Therefore, there is an urgent need for a targeted and improved method and system for pattern recognition network evaluation. Summary of the invention

[0004] The purpose of the present invention is to provide an improved method and system for pattern recognition network evaluation, by establishing a network security parameter set, combining historical data items to form a training sample set, then extracting two samples in sequence, constructing sample pairs, using a branched convolutional neural network to eliminate sample pairs that do not belong to the same category, and re-obtaining a second training sample set, calling a pattern recognition method to train the second sample set, and then solving the optimal solution to obtain the required vector width, and finally establishing an evaluation model, thereby reducing the impact of the number of samples.

[0005] In a first aspect, the present application provides an improved method for pattern recognition network evaluation, the method comprising:

[0006] Collect network environment parameters, aggregate the network environment parameters by item, and establish a network security parameter set;

[0007] Assigning values ​​to the network security parameter set according to the values ​​of the network environment parameters;

[0008] Requesting the server for historical values ​​of the network security parameter set, and listing the historical values ​​and the collected current values ​​in a vector, each item forming a training sample set;

[0009] Extract two samples from the training sample set in sequence to construct a sample pair, input the two samples of the sample pair into the branch convolutional neural network respectively, calculate the similarity probability value between the sample pairs, and judge whether they belong to the same category according to the size of the similarity probability value. If they belong to the same category, their weights are the same and they are stored in the training sample set again. Otherwise, the sample pair is removed from the training sample set;

[0010] After the branched convolutional neural network, a second training sample set is obtained;

[0011] Calling the pattern recognition unit to train the second training sample set, extracting a single historical value in the second training sample set, multiplying the historical value by α, and satisfying a first condition that the sum of the multiplication result and a preset constant a is equal to 0, wherein α is a convex set coefficient, and the reciprocal of the absolute value of α is defined as the vector width; at the same time, subtracting the result of the multiplication of the historical value by α from a preset constant b, and satisfying a second condition that the current value of the single item multiplied by the subtraction result is greater than or equal to 1-β, wherein β is a vector loose coefficient;

[0012] The values ​​of the preset constants a and b depend on the type of the current network. The server pre-stores a mapping relationship between the network type and the preset constants. A and b appear as a pair of constants.

[0013] Calling a computing unit to calculate a partial derivative of the second condition, wherein the partial derivative is calculated based on α and β respectively to obtain an optimal solution for the vector width;

[0014] The optimal solution of the vector width is used as an input parameter of a decision function of a network security assessment model to establish an assessment model;

[0015] The collected network environment parameters are input into the evaluation model to determine whether the network environment is safe.

[0016] In combination with the first aspect, in a first possible implementation of the first aspect, the sub-item aggregation includes a clustering operation, which combines and analyzes local area networks of the same type or adjacent locations, and the sub-item aggregation also includes collecting data according to specified items.

[0017] In combination with the first aspect, in a second possible implementation of the first aspect, when the network environment is determined to be unsafe, the current network environment parameters are recorded and reported to a server.

[0018] In combination with the first aspect, in a third possible implementation of the first aspect, the operation unit adopts a neural network model.

[0019] In a second aspect, the present application provides an improved system for pattern recognition network evaluation, the system comprising a processor and a memory:

[0020] The memory is used to store program code and transmit the program code to the processor;

[0021] The processor is used to execute any one of the four possible methods of the first aspect according to the instructions in the program code.

[0022] In a third aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium is used to store program code, and the program code is used to execute any one of the four possible methods in the first aspect.

[0023] Beneficial Effects

[0024] The present invention provides an improved method and system for pattern recognition network evaluation. By establishing a network security parameter set, combining historical data items to form a training sample set, then sequentially constructing sample pairs, eliminating sample pairs that do not belong to the same category, and re-obtaining a second training sample set, calling a pattern recognition method to train the second sample set, and then solving the optimal solution to obtain the required vector width, and establishing an evaluation model, not only can the dynamic correspondence between scenario parameters and network environment be achieved, but also the influence of the number of samples can be reduced, thereby further improving the efficiency of network security evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0026] Figure 1 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0027] The preferred embodiments of the present invention are described in detail below in conjunction with the accompanying drawings so that the advantages and features of the present invention can be more easily understood by those skilled in the art, thereby making a clearer and more definite definition of the protection scope of the present invention.

[0028] Figure 1 A flowchart of an improved method for pattern recognition network evaluation provided in this application includes:

[0029] Collect network environment parameters, aggregate the network environment parameters by item, and establish a network security parameter set;

[0030] Assigning values ​​to the network security parameter set according to the values ​​of the network environment parameters;

[0031] Requesting the server for historical values ​​of the network security parameter set, and listing the historical values ​​and the collected current values ​​in a vector, each item forming a training sample set;

[0032] Extract two samples from the training sample set in sequence to construct a sample pair, input the two samples of the sample pair into the branch convolutional neural network respectively, calculate the similarity probability value between the sample pairs, and judge whether they belong to the same category according to the size of the similarity probability value. If they belong to the same category, their weights are the same and they are stored in the training sample set again. Otherwise, the sample pair is removed from the training sample set;

[0033] After the branched convolutional neural network, a second training sample set is obtained;

[0034] Calling the pattern recognition unit to train the second training sample set, extracting a single historical value in the second training sample set, multiplying the historical value by α, and satisfying a first condition that the sum of the multiplication result and a preset constant a is equal to 0, wherein α is a convex set coefficient, and the reciprocal of the absolute value of α is defined as the vector width; at the same time, subtracting the result of the multiplication of the historical value by α from a preset constant b, and satisfying a second condition that the current value of the single item multiplied by the subtraction result is greater than or equal to 1-β, wherein β is a vector loose coefficient;

[0035] The values ​​of the preset constants a and b depend on the type of the current network. The server pre-stores a mapping relationship between the network type and the preset constants. A and b appear as a pair of constants.

[0036] Calling a computing unit to calculate a partial derivative of the second condition, wherein the partial derivative is calculated based on α and β respectively to obtain an optimal solution for the vector width;

[0037] The optimal solution of the vector width is used as an input parameter of a decision function of a network security assessment model to establish an assessment model;

[0038] The collected network environment parameters are input into the evaluation model to determine whether the network environment is safe.

[0039] In some preferred embodiments, the sub-item aggregation includes a clustering operation to combine and analyze local area networks of the same type or adjacent locations, and the sub-item aggregation also includes collecting data according to specified items.

[0040] In some preferred embodiments, when the network environment is determined to be unsafe, the current network environment parameters are recorded and reported to a server.

[0041] In some preferred embodiments, the computing unit adopts a neural network model.

[0042] The present application provides an improved system for pattern recognition network evaluation, the system comprising: the system comprising a processor and a memory:

[0043] The memory is used to store program code and transmit the program code to the processor;

[0044] The processor is used to execute the method described in any one of all embodiments of the first aspect according to the instructions in the program code.

[0045] The present application provides a computer-readable storage medium, wherein the computer-readable storage medium is used to store program code, and the program code is used to execute the method described in any one of the embodiments of the first aspect.

[0046] In a specific implementation, the present invention further provides a computer storage medium, wherein the computer storage medium may store a program, and when the program is executed, the program may include some or all of the steps in each embodiment of the present invention. The storage medium may be a disk, an optical disk, a read-only storage memory (abbreviated as: ROM) or a random access memory (abbreviated as: RAM), etc.

[0047] Those skilled in the art can clearly understand that the technology in the embodiments of the present invention can be implemented by means of software plus a necessary general hardware platform. Based on this understanding, the technical solution in the embodiments of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which can be stored in a storage medium such as ROM / RAM, a disk, an optical disk, etc., and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present invention or some parts of the embodiments.

[0048] The same and similar parts between the various embodiments of this specification can be referred to each other. In particular, for the embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the description in the method embodiment.

[0049] The above-described embodiments of the present invention do not limit the protection scope of the present invention.

Claims

1. An improved method for pattern recognition network evaluation, characterized in that: The method comprises: Collect network environment parameters, aggregate the network environment parameters by item, and establish a network security parameter set; Assigning values ​​to the network security parameter set according to the values ​​of the network environment parameters; Requesting the server for historical values ​​of the network security parameter set, and listing the historical values ​​and the collected current values ​​in a vector, each item forming a training sample set; Extract two samples from the training sample set in sequence to construct a sample pair, input the two samples of the sample pair into the branch convolutional neural network respectively, calculate the similarity probability value between the sample pairs, and judge whether they belong to the same category according to the size of the similarity probability value. If they belong to the same category, their weights are the same and they are stored in the training sample set again. Otherwise, the sample pair is removed from the training sample set; After the branched convolutional neural network, a second training sample set is obtained; Calling the pattern recognition unit to train the second training sample set, extracting a single historical value in the second training sample set, multiplying the historical value by α, and satisfying a first condition that the sum of the multiplication result and a preset constant a is equal to 0, wherein α is a convex set coefficient, and the reciprocal of the absolute value of α is defined as the vector width; at the same time, subtracting the result of the multiplication of the historical value by α from a preset constant b, and satisfying a second condition that the current value of the single item multiplied by the subtraction result is greater than or equal to 1-β, wherein β is a vector loose coefficient; The values ​​of the preset constants a and b depend on the type of the current network. The server pre-stores a mapping relationship between the network type and the preset constants. A and b appear as a pair of constants. Calling a computing unit to calculate a partial derivative of the second condition, wherein the partial derivative is calculated based on α and β respectively to obtain an optimal solution for the vector width; The optimal solution of the vector width is used as an input parameter of a decision function of a network security assessment model to establish an assessment model; The collected network environment parameters are input into the evaluation model to determine whether the network environment is safe.

2. The method according to claim 1, characterized in that: The sub-item aggregation includes a clustering operation, which combines and analyzes local area networks of the same type or adjacent locations. The sub-item aggregation also includes collecting data according to specified items.

3. The method according to claim 2, characterized in that: When it is determined that the network environment is unsafe, the current network environment parameters are recorded and reported to the server.

4. The method according to claim 3, characterized in that: The computing unit adopts a neural network model.

5. An improved system for pattern recognition network evaluation, characterized in that: The system comprises a processor and a memory: The memory is used to store program code and transmit the program code to the processor; The processor is used to execute the method according to any one of claims 1 to 4 according to the instructions in the program code.

6. A computer-readable storage medium, characterized in that: The computer-readable storage medium is used to store program codes, and the program codes are used to execute the method according to any one of claims 1 to 4.

Images