A general decision mathematical model system based on unweighted big data objective law

Abstract

The application discloses a general decision mathematical model system based on a non-weighted big data objective law, relates to the technical field of big data processing, artificial intelligence modeling, space-time sequence prediction and intelligent decision system, and comprises the following modules: a full-amount data access and preprocessing module, a non-weighted variable correlation calculation module, an automatic model structure generation module, a multi-dimensional space-time evolution deduction module and a decision result output module. The system realizes cross-scene general intelligent decision without manual intervention and driven by objective data through full-amount data access, non-weighted correlation calculation, automatic model construction, multi-dimensional space-time deduction and equal multi-target decision output, and solves the technical pain points of strong subjectivity, poor universality and single deduction dimension of traditional decision models.

Description

Technical Field

[0001] This invention relates to the fields of big data processing, artificial intelligence modeling, spatiotemporal sequence prediction, and intelligent decision-making systems, and particularly to a general mathematical model system for decision-making based on the objective laws of unweighted big data. Background Technology

[0002] In scenarios such as macroeconomic monitoring, regional development assessment, industrial structure analysis, and public governance decision-making, traditional decision support systems generally suffer from the following technical deficiencies:

[0003] Model construction relies on manually setting indicator weights, which is highly subjective and prone to introducing biases;

[0004] Data filtering relies on empirical rules, making it impossible to automatically include all data in the analysis.

[0005] The model has poor general applicability and needs to be retrained and optimized for different business scenarios.

[0006] Trend extrapolation is mostly a single-dimensional static prediction, which makes it difficult to achieve parallel evolution across multiple spatiotemporal scales;

[0007] The system output is susceptible to human intervention and lacks a stable, reproducible, and objective decision-making output mechanism.

[0008] Therefore, in order to address the problem that existing big data analysis and machine learning models focus on optimization of specific scenarios and cannot achieve objective extrapolation across multiple spatiotemporal scales under a unified architecture without manual assignment, subjective screening, or remodeling, resulting in low efficiency and insufficient prediction accuracy of decision-making systems when migrating across scenarios, a general decision-making mathematical model system based on the objective laws of unweighted big data is needed. Summary of the Invention

[0009] The present invention aims to at least partially solve one of the technical problems in the above-mentioned technologies.

[0010] To achieve the above objectives, the first aspect of this invention proposes a general decision-making mathematical model system based on the objective laws of unweighted big data, comprising: a full data access and preprocessing module, an unweighted variable correlation calculation module, an automatic model structure generation module, a multi-dimensional spatiotemporal evolution deduction module, and a decision result output module; the system runs on a computer hardware cluster, performs data processing and model deduction through a distributed computing framework, and does not introduce manual weight assignment, preset indicator priorities, or perform subjective data screening throughout the entire process, but is driven by the inherent statistical laws of the data to generate and evolve the model.

[0011] In addition, the general decision-making mathematical model system based on the objective laws of unweighted big data proposed above according to the present invention may also have the following additional technical features:

[0012] Furthermore, the full data access and preprocessing module performs standardization processing on the data: It is used to eliminate the influence of dimensions, retain all valid indicators, and does not perform subjective rejection or sample pruning.

[0013] Furthermore, the unweighted variable association calculation module uses mutual information to measure variable association: The strength of the association between variables is automatically generated by the data distribution, without any manual weighting intervention.

[0014] Furthermore, the automatic model structure generation module enables self-construction of model topology, self-discovery of causal paths, self-calibration of parameters, and self-optimization of goodness of fit.

[0015] Furthermore, the multidimensional spatiotemporal evolution deduction module performs state evolution in parallel according to short-term, medium-term, and long-term time scales and local, regional, and global spatial scales: Automatically identifies trend inflection points, critical points, and risk thresholds.

[0016] Furthermore, the decision result output module outputs a decision scheme based on equal multi-objective optimization: Output objective prediction results, risk warnings and assessment reports.

[0017] Furthermore, it is applicable to scenarios such as macroeconomic monitoring, regional development analysis, industrial structure assessment, pharmaceutical industry operation analysis, and public policy evaluation.

[0018] Furthermore, it supports policy simulation comparison, historical effect backtesting, risk warning, and continuous model iteration and optimization.

[0019] According to the present invention, a general decision-making mathematical model system based on the objective laws of unweighted big data achieves cross-domain, non-interventional, and highly stable objective decision output through unbiased processing of all data, automatic calculation of variable correlation, autonomous generation of model structure, multi-temporal parallel inference and equal multi-objective decision-making, thereby significantly improving the universality, objectivity and inference accuracy of the decision-making model. Attached Figure Description

[0020] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0021] Figure 1 This is an overall structural block diagram of a general decision mathematical model system based on the objective laws of unweighted big data according to an embodiment of the present invention;

[0022] Figure 2 This is an unweighted automatic modeling flowchart of a general decision mathematical model system based on the objective laws of unweighted big data according to an embodiment of the present invention;

[0023] Figure 3 This is a schematic block diagram of the multidimensional spatiotemporal evolution deduction logic of a general decision mathematical model system based on the objective laws of unweighted big data according to an embodiment of the present invention:

[0024] As shown in the figure:

[0025] 1. Full data access and preprocessing module; 2. Unweighted variable correlation calculation module; 3. Automatic model structure generation module; 4. Multidimensional spatiotemporal evolution inference module; 5. Decision result output module. Detailed Implementation

[0026] Embodiments of the present invention are described in detail below, with examples of the embodiments illustrated in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0027] The following description, in conjunction with the accompanying drawings, illustrates a general decision-making mathematical model system based on the objective laws of unweighted big data, according to an embodiment of the present invention.

[0028] like Figures 1 to 3 As shown in the figure, a general decision mathematical model system based on the objective laws of unweighted big data according to an embodiment of the present invention includes: a full data access and preprocessing module, an unweighted variable correlation calculation module, an automatic model structure generation module, a multi-dimensional spatiotemporal evolution deduction module, and a decision result output module. The system runs on a computer hardware cluster and performs data processing and model deduction through a distributed computing framework. The entire process does not introduce manual weight assignment, does not preset indicator priorities, and does not perform subjective data screening. The model generation and evolution are driven by the inherent statistical laws of the data.

[0029] It should be noted that the full data access and preprocessing module is used to collect multi-source heterogeneous time-series data, including but not limited to economic indicators, industry data, population structure, resource distribution, spatial geography, market operation, and historical time-series data. It performs missing value imputation, outlier detection, standardization, and normalization on the raw data. ,in For the original indicators, The mean, The standard deviation is used to standardize the data scale, ensuring that all valid indicators participate equally in subsequent calculations, without performing subjective rejection or sample pruning.

[0030] It should be noted that the unweighted variable association calculation module includes all standardized indicators as equal variables in the computational space, and measures the strength of the association between variables through statistical correlation and mutual information: The association weights of variables are generated automatically by the data distribution, without setting manual prior weights or introducing expert experience corrections, thus forming an unbiased association matrix.

[0031] It should be noted that the automatic model structure generation module is based on the unbiased correlation matrix and automatically performs: variable hierarchical clustering and causal path mining, self-construction of model topology, self-calibration of model order and parameters, and iterative optimization of goodness of fit. The automatic model structure generation module adopts a combination of forward stepwise construction and backward pruning strategies to form a stable and inferable mathematical model without the need for manual intervention in the modeling process.

[0032] It should be noted that the multidimensional spatiotemporal evolution deduction module constructs a nested deduction framework of time dimension (short-term T1, medium-term T2, long-term T3) and spatial dimension (local unit, regional set, global system): ,in Let be the system state vector. It is an unbiased incidence matrix. As an evolution operator, the multi-dimensional spatiotemporal evolution simulation module executes multi-scenario simulations in parallel, automatically identifying trend inflection points, critical points, risk thresholds, and periodic characteristics.

[0033] It should be noted that the decision output module is based on multi-temporal and spatial simulation results, and solves for the optimal decision path through multi-objective equal optimization: It outputs objective decision-making solutions, risk warning information, trend prediction curves, and effect evaluation reports without modifying the results or adjusting the stance.

[0034] Specifically, firstly, multi-source time-series data from the target domain are collected to construct a unified standardized dataset. Subsequently, the data undergoes preprocessing such as missing value imputation, outlier removal, standardization, and normalization to ensure data quality and computational consistency. Secondly, an unweighted correlation matrix of all variables is constructed, and mutual information and statistical correlation are calculated based on the inherent distribution of the data. The model topology is automatically generated, and parameter calibration and iterative model optimization are achieved. Thirdly, parallel evolutionary deduction in multiple spatiotemporal dimensions is initiated to accurately identify development trends, inflection point characteristics, and risk intervals. Based on a multi-objective equal optimization mechanism, decision-making schemes, evaluation reports, and early warning information are output. Finally, historical backtesting, policy simulation comparison, and continuous iterative updates throughout the model's entire lifecycle are supported.

[0035] Example 1: Regional Economic Operation Analysis System: Collects full-volume time-series data such as regional GDP, industrial added value, fixed asset investment, resident consumption, population flow, and innovation indicators. After standardization, the data is used for unweighted correlation calculation. The system automatically constructs an economic operation evolution model, performs short-term, medium-term, and long-term projections, and outputs industrial structure optimization paths, investment allocation suggestions, risk warning ranges, and development trend predictions.

[0036] Example 2: Pharmaceutical Industry Operation Decision Analysis: By accessing data such as pharmaceutical R&D investment, clinical trial progress, production capacity, supply chain indicators, market coverage, medical insurance payment, and policy constraints, the system automatically mines variable correlations without weighting, constructs an industry evolution model, identifies R&D transformation breakpoints, weak links in the supply chain, and market supply and demand equilibrium points, and outputs industry optimization strategies and risk prevention and control solutions.

[0037] In summary, the general decision-making mathematical model system based on the objective laws of unweighted big data according to the embodiments of the present invention completely abandons manual weighting and subjective screening, takes the inherent laws of data as the core driving force, realizes automatic model construction and accurate multi-temporal and spatial extrapolation, and can be widely adapted to the decision-making needs of multiple fields such as macroeconomics, industrial development, and public policy. It has the core advantages of high versatility, high objectivity and high stability.

[0038] In the description of this specification, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0039] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0040] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A general decision-making mathematical model system based on the objective laws of unweighted big data, characterized in that, include: The module includes a full data access and preprocessing module (1), an unweighted variable association calculation module (2), an automatic model structure generation module (3), a multidimensional spatiotemporal evolution deduction module (4), and a decision result output module (5). The system is based on distributed execution of computer clusters. It does not introduce manual weight assignment, preset indicator priorities, or perform subjective data screening. The model generation and evolution are driven by the inherent statistical laws of the data, which is used to improve the objectivity of decision output and the accuracy of inference.

2. The system according to claim 1, characterized in that, The full data access and preprocessing module (1) performs standardization processing on the data: ； This standardization is used to eliminate differences in the dimensions of indicators, retain all valid indicators to participate in the calculation equally, and avoid subjective elimination and sample pruning.

3. The system according to claim 1, characterized in that, The unweighted variable association calculation module (2) associates variables through mutual information measurement: ； The strength of the association between variables is automatically generated by the data distribution, without human intervention in weighting, thus avoiding calculation bias caused by prior experience.

4. The system according to claim 1, characterized in that, The automatic model structure generation module (3) realizes self-construction of model topology, self-discovery of causal paths, self-calibration of parameters and self-optimization of fit.

5. The system according to claim 1, characterized in that, The multidimensional spatiotemporal evolution deduction module (4) performs state evolution in parallel according to short-term, medium-term, and long-term time scales and local, regional, and global spatial scales: ； Used to automatically identify trend inflection points, system critical points, and risk thresholds.

6. The system according to claim 1, characterized in that, The decision result output module (5) outputs a decision scheme based on equal multi-objective optimization: ； Output objective forecast results, risk warnings and assessment reports, without any bias or artificial adjustment.

7. The system according to claim 1, characterized in that, It is applicable to scenarios such as macroeconomic monitoring, regional development analysis, industrial structure assessment, pharmaceutical industry operation analysis, and public policy evaluation.

8. The system according to claim 1, characterized in that, It supports policy simulation comparison, historical effect backtesting, risk warning and continuous model iteration and optimization.

Images