AI-based data analysis method and terminal

Abstract

The invention discloses a data analysis method based on AI, and the method comprises the steps: building a plurality of calculation class blocks for each calculation scene, receiving a mathematical formula model of the calculation scene inputted by a user, and filling the mathematical formula model into the calculation class blocks; obtaining training data of the calculation scene, matching the training data with the mathematical formula models of the calculation class blocks according to a matching algorithm, and performing relation derivation through AI to obtain a mathematical association relation between the calculation class blocks; filling a mathematical data model library with the calculation class blocks and the mathematical association relationship; acquiring to-be-processed data input by a user, performing matching through the mathematical data model library to obtain the calculation scene, the corresponding calculation class block and the mathematical association relationship, and then performing mathematical calculation; a mathematical data model of a specific scene can be established in a more targeted manner, the problem of dirty data is solved, and analysis and calculation can be performed more accurately.

Description

Technical Field

[0001] The present invention relates to the field of technical analysis technology, and in particular to an AI-based data analysis method and terminal. Background Art

[0002] As we all know, AI requires large models for training, hardware support, and a large amount of natural language and other types of data, so dirty data is inevitable. Because there are too many data samples and complex design scenarios, it is not completed for a single scenario, so there may be a lot of logic that leads to errors. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide an AI-based data analysis method and terminal that are applicable to multivariate formulas in different computing scenarios while retaining the simplicity and complexity of each computing scenario.

[0004] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0005] An AI-based data analysis method comprises the following steps:

[0006] S1. For each computing scenario, establish multiple computing blocks, receive the mathematical formula model of the computing scenario input by the user and fill it into the computing block;

[0007] S2, obtaining training data of the computing scenario, matching the training data with the mathematical formula model of the computing class block according to a matching algorithm, and performing relationship deduction through AI to obtain a mathematical correlation relationship between the computing class blocks;

[0008] S3, filling the calculation class block and the mathematical association relationship into a mathematical data model library;

[0009] S4. Obtain the data to be processed input by the user, match it through the mathematical data model library, obtain the calculation scenario and the corresponding calculation class block and the mathematical association relationship, and then perform mathematical calculation.

[0010] In order to solve the above technical problems, another technical solution adopted by the present invention is:

[0011] An AI-based data analysis terminal includes a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program:

[0012] S1. For each computing scenario, establish multiple computing blocks, receive the mathematical formula model of the computing scenario input by the user and fill it into the computing block;

[0013] S2, obtaining training data of the computing scenario, matching the training data with the mathematical formula model of the computing class block according to a matching algorithm, and performing relationship deduction through AI to obtain a mathematical correlation relationship between the computing class blocks;

[0014] S3, filling the calculation class block and the mathematical association relationship into a mathematical data model library;

[0015] S4. Obtain the data to be processed input by the user, match it through the mathematical data model library, obtain the calculation scenario and the corresponding calculation class block and the mathematical association relationship, and then perform mathematical calculation.

[0016] The beneficial effects of the present invention are as follows: an AI-based data analysis method and terminal of the present invention, for different computing scenarios, fills mathematical formula models into multiple computing class blocks of the scenario, and derives the mathematical association relationship between different computing class blocks by combining training data with the mathematical formula model, so as to establish a mathematical data model for a specific scenario in a more targeted manner, solve the dirty data problem, and determine the computing scenario of the data to be processed input by the user by matching it with the mathematical data model, thereby being able to perform analysis and calculation more accurately. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a flow chart of an AI-based data analysis method according to an embodiment of the present invention;

[0018] Figure 2 This is a structural diagram of an AI-based data analysis terminal according to an embodiment of the present invention;

[0019] Figure 3 This is a schematic diagram of an abstract architecture of an AI-based data analysis method according to an embodiment of the present invention;

[0020] Figure 4 Schematic diagram of an AI-based data analysis method according to an embodiment of the present invention;

[0021] Description of labels:

[0022] 1. An AI-based data analysis terminal; 2. Processor; 3. Memory. DETAILED DESCRIPTION

[0023] In order to explain the technical content, achieved objectives and effects of the present invention in detail, the following is an explanation in combination with the implementation modes and the accompanying drawings.

[0024] Please refer to Figure 1 , an AI-based data analysis method, comprising the steps of:

[0025] S1. For each computing scenario, establish multiple computing blocks, receive the mathematical formula model of the computing scenario input by the user and fill it into the computing block;

[0026] S2, obtaining training data of the computing scenario, matching the training data with the mathematical formula model of the computing class block according to a matching algorithm, and performing relationship deduction through AI to obtain a mathematical correlation relationship between the computing class blocks;

[0027] S3, filling the calculation class block and the mathematical association relationship into a mathematical data model library;

[0028] S4. Obtain the data to be processed input by the user, match it through the mathematical data model library, obtain the calculation scenario and the corresponding calculation class block and the mathematical association relationship, and then perform mathematical calculation.

[0029] From the above description, it can be seen that the beneficial effects of the present invention are: an AI-based data analysis method and terminal of the present invention, for different computing scenarios, fills mathematical formula models into multiple computing class blocks of the scene, and derives the mathematical correlation relationship between different computing class blocks by combining training data with mathematical formula models, which can more specifically establish mathematical data models for specific scenarios and solve the dirty data problem. For the data to be processed input by the user, the computing scenario of the data to be processed is determined by matching it with the mathematical data model, so that more accurate analysis and calculation can be performed.

[0030] Furthermore, the matching algorithm includes a dynamic programming algorithm, a divide-and-conquer algorithm, a greedy algorithm, and a random forest regression algorithm.

[0031] From the above description, it can be seen that the matching algorithm used can be a dynamic programming algorithm, a divide-and-conquer algorithm, a greedy algorithm, a random forest regression algorithm, etc., which can more accurately adapt the training data to each computing block and more accurately derive the mathematical correlation between each computing block.

[0032] Furthermore, the mathematical association relationship refers to a calculation formula model for calculating the data content of another calculation block based on the data content of any two or more calculation blocks in the scenario.

[0033] From the above description, it can be seen that through mathematical association relationships, data between different computing class blocks can be mutually derived.

[0034] Furthermore, in each of the computing scenarios, the number of the computing blocks is 4, and the mathematical association relationship refers to a calculation formula model for calculating the data content of another computing block based on the data content of any three of the computing blocks.

[0035] From the above description, it can be seen that the number of calculation blocks is defined as 4, which avoids too many calculation blocks causing complicated mathematical associations while ensuring their effectiveness.

[0036] Please refer to Figure 2 , an AI-based data analysis terminal, comprising a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program:

[0037] S1. For each computing scenario, establish multiple computing blocks, receive the mathematical formula model of the computing scenario input by the user and fill it into the computing block;

[0038] S2, obtaining training data of the computing scenario, matching the training data with the mathematical formula model of the computing class block according to a matching algorithm, and performing relationship deduction through AI to obtain a mathematical correlation relationship between the computing class blocks;

[0039] S3, filling the calculation class block and the mathematical association relationship into a mathematical data model library;

[0040] S4. Obtain the data to be processed input by the user, match it through the mathematical data model library, obtain the calculation scenario and the corresponding calculation class block and the mathematical association relationship, and then perform mathematical calculation.

[0041] From the above description, it can be seen that the beneficial effects of the present invention are: an AI-based data analysis method and terminal of the present invention, for different computing scenarios, fills mathematical formula models into multiple computing class blocks of the scene, and derives the mathematical correlation relationship between different computing class blocks by combining training data with mathematical formula models, which can more specifically establish mathematical data models for specific scenarios and solve the dirty data problem. For the data to be processed input by the user, the computing scenario of the data to be processed is determined by matching it with the mathematical data model, so that more accurate analysis and calculation can be performed.

[0042] Furthermore, the matching algorithm includes a dynamic programming algorithm, a divide-and-conquer algorithm, a greedy algorithm, and a random forest regression algorithm.

[0043] From the above description, it can be seen that the matching algorithm used can be a dynamic programming algorithm, a divide-and-conquer algorithm, a greedy algorithm, a random forest regression algorithm, etc., which can more accurately adapt the training data to each computing block and more accurately derive the mathematical correlation between each computing block.

[0044] Furthermore, the mathematical association relationship refers to a calculation formula model for calculating the data content of another calculation block based on the data content of any two or more calculation blocks in the scenario.

[0045] From the above description, it can be seen that through mathematical association relationships, data between different computing class blocks can be mutually derived.

[0046] Furthermore, in each of the computing scenarios, the number of the computing blocks is 4, and the mathematical association relationship refers to a calculation formula model for calculating the data content of another computing block based on the data content of any three of the computing blocks.

[0047] From the above description, it can be seen that the number of calculation blocks is defined as 4, which avoids too many calculation blocks causing complicated mathematical associations while ensuring their effectiveness.

[0048] Please refer to Figure 1 , Figure 3 and Figure 4 , Embodiment 1 of the present invention is:

[0049] An AI-based data analysis method comprises the following steps:

[0050] S1. For each computing scenario, establish multiple computing blocks, receive the mathematical formula model of the computing scenario input by the user and fill it into the computing block;

[0051] In this embodiment, in each of the computing scenarios, the number of the computing class blocks is 4.

[0052] In this embodiment, reference can be made to Figure 3 For each computing scenario, the algorithm is divided into four computing blocks, which can be anthropomorphically classified into "personality", "behavior", "mood" and "spirit" blocks. Mathematical formula models are assigned to the computing blocks.

[0053] Please refer to Figure 4 , reasonably assign mathematical formula models (formulas / models corresponding to associated data) to the calculation blocks involved in the retrospective summary, and enter mathematical formulas that meet the conditions. Pre-enumerate the four most important components of the scene (there may be more than 4, or less than 4), and temporarily discard some of them or split them into the fourth part. Through data research, temporarily discard the relationship with the other four parts, for example:

[0054] Scenario: Summarized as: S scenario, composed of: s1, s2, s3, s4, s5

[0055] S=s1+s2+s3+s4

[0056] s1=a0+t0+n0

[0057] s2=f0+g0+a0

[0058] s3=a0+n0+f0

[0059] s4=f1+g2

[0060] s5=f0+a0+t0

[0061] As above: S scene is composed of S0-S5. At this time, it is uncertain how s0-s5 are related to scene S, so it needs to be discussed. Through discussion, it is found that the composition formula of s4 is not closely related to s1-s3, while the composition of s5 is related to s1 s2 s3. The composition formula of s4 has nothing to do with s1 s2 s3 s5. If it is added to the conditions, no restriction will be generated. The conclusion is that s4 is not closely related to the S scene, so the s4 association formula is removed and replaced and s5 is added.

[0062] S2, obtaining training data of the computing scenario, matching the training data with the mathematical formula model of the computing class block according to a matching algorithm, and performing relationship deduction through AI to obtain a mathematical correlation relationship between the computing class blocks;

[0063] The matching algorithms include dynamic programming algorithm, divide-and-conquer algorithm, greedy algorithm and random forest regression algorithm;

[0064] The mathematical association relationship refers to a calculation formula model for calculating the data content of another calculation block based on the data content of any two or more calculation blocks in the scenario;

[0065] In this embodiment, the mathematical association relationship refers to a calculation formula model for calculating the data content of another calculation block based on the data content of any three of the calculation blocks.

[0066] In this embodiment, the data source (training data) is input into the mathematical formula involved in the algorithm so that the data source can match the corresponding mathematical formula. Specifically, the data source conditions are matched through "dynamic programming", "divide and conquer algorithm", "greedy algorithm" and "random forest regression algorithm".

[0067] S3, filling the calculation class block and the mathematical association relationship into a mathematical data model library;

[0068] S4. Obtain the data to be processed input by the user, match it through the mathematical data model library, obtain the calculation scenario and the corresponding calculation class block and the mathematical association relationship, and then perform mathematical calculation.

[0069] For uncertain data samples (data to be processed input by users), the model formulas in unknown scenarios are summarized by backtracking, and the corresponding scenario formulas are matched and inferred for processing.

[0070] For example, take the stock market trading as a dirty data sample, and set the target of 5% return rate for half a year. Take the "personality" in the regression summary as the mathematical formula model of "transaction amount" as the attribute-related matching, "behavior" as the degree of increase as the floating model, "mood" rise or fall as the model, and "spirit" as the transaction amount. After confirming the attributes, we can explore: the "mood" model relationship caused by transaction volume × transaction price = transaction amount. Then we can easily get the "random forest regression algorithm" used above.

[0071] Please refer to Figure 2 , Embodiment 2 of the present invention is:

[0072] An AI-based data analysis terminal 1 includes a processor 2, a memory 3, and a computer program stored in the memory 3 and executable on the processor 2. When the processor 2 executes the computer program, the steps of the AI-based data analysis method in the above embodiment 1 are implemented.

[0073] To summarize, the present invention provides an AI-based data analysis method and terminal, which, for different computing scenarios, fills mathematical formula models into multiple computing blocks of the scenario, and derives the mathematical correlation between different computing blocks by combining training data with the mathematical formula model. This enables a more targeted mathematical data model for a specific scenario to solve the dirty data problem, and for the data to be processed input by the user, the computing scenario of the data to be processed is determined by matching it with the mathematical data model, thereby enabling more accurate analysis and calculation.

[0074] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent transformations made using the contents of the present invention's specification and drawings, or directly or indirectly applied in related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A data analysis method based on AI, characterized in that: Includes steps: S1. For each computing scenario, establish multiple computing blocks, receive the mathematical formula model of the computing scenario input by the user and fill it into the computing block; S2, obtaining training data of the computing scenario, matching the training data with the mathematical formula model of the computing class block according to a matching algorithm, and performing relationship deduction through AI to obtain a mathematical correlation relationship between the computing class blocks; S3, filling the calculation class block and the mathematical association relationship into a mathematical data model library; S4. Obtain the data to be processed input by the user, match it through the mathematical data model library, obtain the calculation scenario and the corresponding calculation class block and the mathematical association relationship, and then perform mathematical calculation.

2. The AI-based data analysis method according to claim 1, characterized in that: The matching algorithms include dynamic programming algorithm, divide-and-conquer algorithm, greedy algorithm and random forest regression algorithm.

3. The AI-based data analysis method according to claim 1, characterized in that: The mathematical association relationship refers to a calculation formula model for calculating the data content of another calculation block based on the data content of any two or more calculation blocks in the scenario.

4. The AI-based data analysis method according to claim 3, characterized in that: In each of the computing scenarios, the number of the computing blocks is 4, and the mathematical association relationship refers to a computing formula model for calculating the data content of another computing block based on the data content of any three of the computing blocks.

5. An AI-based data analysis terminal, comprising a processor, a memory, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the following steps are implemented: S1. For each computing scenario, establish multiple computing blocks, receive the mathematical formula model of the computing scenario input by the user and fill it into the computing block; S2, obtaining training data of the computing scenario, matching the training data with the mathematical formula model of the computing class block according to a matching algorithm, and performing relationship deduction through AI to obtain a mathematical correlation relationship between the computing class blocks; S3, filling the calculation class block and the mathematical association relationship into a mathematical data model library; S4. Obtain the data to be processed input by the user, match it through the mathematical data model library, obtain the calculation scenario and the corresponding calculation class block and the mathematical association relationship, and then perform mathematical calculation.

6. The AI-based data analysis terminal according to claim 5, characterized in that: The matching algorithms include dynamic programming algorithm, divide-and-conquer algorithm, greedy algorithm and random forest regression algorithm.

7. The AI-based data analysis terminal according to claim 5, characterized in that: The mathematical association relationship refers to a calculation formula model for calculating the data content of another calculation block based on the data content of any two or more calculation blocks in the scenario.

8. The AI-based data analysis terminal according to claim 7, characterized in that: In each of the computing scenarios, the number of the computing blocks is 4, and the mathematical association relationship refers to a computing formula model for calculating the data content of another computing block based on the data content of any three of the computing blocks.

Images