Prediction algorithm for predicting international crude oil price
Through the multivariate regression analysis method combined with multidimensional data and key influencing factors, an international crude oil price prediction model was constructed, which solved the problem of low prediction accuracy in the existing technology and achieved higher prediction accuracy and stability.
Patent Information
- Application Number
- CN202411792539.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-07
- Publication Date
- 2025-03-28
AI Technical Summary
The prior art has problems of poor generalization ability and low prediction accuracy in international crude oil price prediction.
Multivariate regression analysis method is used to combine multi-dimensional data and key influencing factors to build an efficient prediction model. Specific steps include data collection and processing, construction of multiple regression equations, model verification and adjustment, and real-time monitoring and prediction.
It improves the accuracy and stability of international crude oil price prediction, provides decision makers with scientific and accurate decision-making basis, and is highly practical and operable.
Smart Images

Figure BDA0005175682310000021 
Figure BDA0005175682310000041 
Figure FDA0005175682300000011
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of prediction methods, and in particular to a prediction algorithm for predicting international crude oil prices. Background Art
[0002] With the continuous development of the global economy and the growing demand for energy, the fluctuation of international crude oil prices has a profound impact on the economy, industry and people's livelihood of various countries. Therefore, it is of great strategic significance to accurately predict the trend of international crude oil prices.
[0003] Existing technologies mostly use single time dimension data or a single influencing factor for modeling. Therefore, the existing technologies have the problem that the prediction models involved have poor generalization ability and the prediction values output by the models are not accurate, which is not effective for predicting international crude oil prices. Summary of the invention
[0004] In view of the above background technology, those skilled in the art provide a prediction algorithm for predicting international crude oil prices to solve the problem of low accuracy in predicting international crude oil prices in the prior art.
[0005] The present invention provides a prediction algorithm for predicting international crude oil prices, the composition scheme is as follows, including:
[0006] S1 Data collection and processing: Obtain multi-dimensional data related to international crude oil prices, clean the data and perform standardization;
[0007] S2 builds a multivariate model: based on the collected data, selects appropriate independent variable factors (influencing factors brought by multi-dimensional data) and dependent variable factors (international crude oil prices), and builds a multivariate regression equation; uses statistical software to fit and optimize the model to ensure the prediction accuracy and stability of the model;
[0008] S3 Model Verification and Adjustment: Evaluate the model's prediction performance by comparing the model's prediction results with the actual crude oil prices; make necessary adjustments and optimizations to the model based on the evaluation results to improve the prediction accuracy;
[0009] S4 Real-time monitoring and forecasting: Regularly monitor the independent variable factors in real time, input the latest data into the forecasting model, and obtain the forecast results of crude oil prices in the future.
[0010] Preferably, the multi-dimensional data in step S1 is based on multiple influencing factors, including but not limited to crude oil market price, US dollar exchange rate, global economic growth rate, oil production, geopolitical risk index, supply and demand relationship, etc.
[0011] Preferably, the data standardization process in step S1 includes the following steps: the data is standardized according to the following formula:
[0012]
[0013] Among them, ξ i is the data value after the standardized transformation, x i is the verification data after the unit root verification of the data, is the mean value of the verification data in this data column, and σ is the standard deviation of the verification data in this data column.
[0014] Preferably, step S1 specifically includes: cleaning the data, filling missing values, removing outliers and duplicate data, and standardizing the data so that the model can learn and predict better. For outliers, they are removed by setting a suitable threshold; for missing values, they are filled by upsampling.
[0015] Preferably: the upsampling method specifically comprises the following steps:
[0016] S101: Divide the missing time period into n equal parts according to the minimum spacing of time points in the non-missing data time period, and obtain n-1 time points in the missing time period;
[0017] S102: At n-1 time points, use the formula
[0018] 1≤m≤n-1
[0019] Get the data value P corresponding to the mth time point in the missing time period m , where a0 is the data value corresponding to the last time point before the missing time period, a n is the data value corresponding to the first time point after the missing period.
[0020] Preferably, the multivariate regression equation in step S2 is:
[0021] Y t =β0+β1×X1+β2×X2+β3×X3+β4×X4+β5×X5+ε
[0022] Among them, Y t The dependent variable is the international crude oil price in the tth period; β0 is the intercept term; β1, β2, β3, β4, and β5 are the regression coefficients of the influencing factors: US dollar exchange rate, global economic growth rate, oil production, geopolitical risk index, and supply and demand relationship, indicating the degree of impact on the international crude oil price when each influencing factor changes by one unit; ε is the error term, representing the part that cannot be explained in the model. ε is the difference between the actual result and the predicted result, which is included in the model to explain this subtle change.
[0023] Utilizing the collected historical data, which have been standardized, statistical software was used to perform regression analysis on each influencing factor to obtain the regression model coefficients of β1, β2, β3, β4, and β5.
[0024] Preferably, the statistical software includes using Python's statsmode ls library or R language to perform regression analysis, and using regression analysis to calculate the estimated value of each regression coefficient.
[0025] Preferably, the evaluation model in step S3 is built using open source Hadoop ecosystem components and Python programs.
[0026] In further explanation, the forecast results will be released to the government, enterprises and investors through official websites, research reports, etc., to provide strong support for their decision-making. At the same time, corresponding response strategies will be formulated based on the forecast results to deal with the risks brought by the fluctuation of crude oil market prices. By real-time monitoring of key influencing factors, accurate prediction of crude oil price trends can be achieved.
[0027] Technical effects and advantages of the present invention:
[0028] The present invention adopts multiple regression analysis method and combines the key factors affecting crude oil prices to build an efficient and practical prediction model, providing scientific and accurate decision-making basis for decision makers. At the same time, the scheme has high practicality and operability, and is expected to provide strong support for governments, enterprises and investors, and promote the stable development of the energy market. DETAILED DESCRIPTION
[0029] The present invention will be further described in detail below in conjunction with specific embodiments. The embodiments of the present invention are provided for the purpose of illustration and description, and are not intended to be exhaustive or to limit the present invention to the disclosed forms. Many modifications and variations will be apparent to those of ordinary skill in the art. The embodiments are selected and described to better illustrate the principles and practical applications of the present invention, and to enable those of ordinary skill in the art to understand the present invention and thereby design various embodiments with various modifications suitable for specific uses.
[0030] Example 1
[0031] In this embodiment, a prediction algorithm for predicting international crude oil prices is provided, including:
[0032] S0 builds a data platform for predicting international crude oil prices. The data platform is used in the following steps to collect and process data;
[0033] S1 Data collection and processing: Obtain multi-dimensional data related to international crude oil prices, clean the data and perform standardization;
[0034] The multi-dimensional data in step S1 is based on multiple influencing factors, including but not limited to crude oil market price, US dollar exchange rate, global economic growth rate, oil production, geopolitical risk index, supply and demand relationship, etc.
[0035] Preferably, the data standardization process in step S1 includes the following steps: the data is standardized according to the following formula:
[0036]
[0037] Among them, ξ i is the data value after the standardized transformation, x i is the verification data after the unit root verification of the data, is the mean value of the verification data in this data column, and σ is the standard deviation of the verification data in this data column.
[0038] Preferably, step S1 specifically includes: cleaning the data, filling missing values, removing outliers and duplicate data, and standardizing the data so that the model can learn and predict better. For outliers, they are removed by setting a suitable threshold; for missing values, they are filled by upsampling.
[0039] Preferably: the upsampling method specifically comprises the following steps:
[0040] S101: Divide the missing time period into n equal parts according to the minimum spacing of time points in the non-missing data time period, and obtain n-1 time points in the missing time period;
[0041] S102: At n-1 time points, use the formula
[0042] 1≤m≤n-1
[0043] Get the data value P corresponding to the mth time point in the missing time period m , where a0 is the data value corresponding to the last time point before the missing time period, a n is the data value corresponding to the first time point after the missing period.
[0044] S2 builds a multivariate model: Based on the collected data, select appropriate independent variable factors (influencing factors brought by multi-dimensional data) and dependent variable factors (international crude oil prices) to construct a multivariate regression equation; use statistical software to fit and optimize the model to ensure the prediction accuracy and stability of the model.
[0045] Preferably, the multivariate regression equation in step S2 is:
[0046] Y t=β0+β1×X1+β2×X2+β3×X3+β4×X4+β5×X5+ε
[0047] Among them, Y t The dependent variable is the international crude oil price in the tth period; β0 is the intercept term; β1, β2, β3, β4, and β5 are the regression coefficients of the influencing factors: US dollar exchange rate, global economic growth rate, oil production, geopolitical risk index, and supply and demand relationship, indicating the degree of impact on the international crude oil price when each influencing factor changes by one unit; ε is the error term, representing the part that cannot be explained in the model. ε is the difference between the actual result and the predicted result, which is included in the model to explain this subtle change.
[0048] Utilizing the collected historical data, which have been standardized, statistical software was used to perform regression analysis on each influencing factor to obtain the regression model coefficients of β1, β2, β3, β4, and β5.
[0049] Preferably, the statistical software includes using Python's statsmode ls library or R language to perform regression analysis, and using regression analysis to calculate the estimated value of each regression coefficient.
[0050] S3 Model Verification and Adjustment: Evaluate the model’s prediction performance by comparing the model’s prediction results with the actual crude oil prices; make necessary adjustments and optimizations to the model based on the evaluation results to improve the prediction accuracy.
[0051] The evaluation model in step S3 is built using open source Hadoop ecosystem components and Python programs.
[0052] S4 Real-time monitoring and forecasting: Regularly monitor the independent variable factors in real time, input the latest data into the forecasting model, and obtain the forecast results of crude oil prices in the future.
[0053] In further explanation, the forecast results will be released to the government, enterprises and investors through official websites, research reports, etc., to provide strong support for their decision-making. At the same time, corresponding response strategies will be formulated based on the forecast results to deal with the risks brought by the fluctuation of crude oil market prices. By real-time monitoring of key influencing factors, accurate prediction of crude oil price trends can be achieved.
[0054] Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field and related fields without creative work should fall within the scope of protection of the present invention. The structures, devices and operating methods not specifically described and explained in the present invention are implemented according to the conventional means in the field unless otherwise specified and limited.
Claims
1. A prediction algorithm for predicting international crude oil prices, characterized in that: include: S1 Data collection and processing: Obtain multi-dimensional data related to international crude oil prices, clean the data and perform standardization; S2 builds a multivariate model: Based on the collected data, selects appropriate multidimensional data as independent variable factors, selects international crude oil prices as dependent variable factors, and builds a multivariate regression equation; S3 Model Verification and Adjustment: Evaluate the model's prediction performance by comparing the model's prediction results with the actual crude oil prices; make necessary adjustments and optimizations to the model based on the evaluation results to improve the prediction accuracy; S4 Real-time monitoring and forecasting: Regularly monitor the independent variable factors in real time, input the latest data into the forecasting model, and obtain the forecast results of crude oil prices in the future.
2. A prediction algorithm for predicting international crude oil prices according to claim 1, characterized in that: The multi-dimensional data in step S1 is based on multiple influencing factors, including the US dollar exchange rate, global economic growth rate, oil production, geopolitical risk index and supply and demand relationship.
3. A prediction algorithm for predicting international crude oil prices according to claim 2, characterized in that: The data standardization process in step S1 includes the following steps: the data is standardized according to the following formula: In formula (1), ξ i is the data value after standardized transformation, x i is the verification data after the unit root verification of the data, is the mean value of the verification data in this data column, and σ is the standard deviation of the verification data in this data column.
4. A prediction algorithm for predicting international crude oil prices according to claim 3, characterized in that: The step S1 specifically includes: cleaning the data, filling missing values, removing outliers and duplicate data, and standardizing the data so that the model can better learn and predict; for outliers, removing them by setting appropriate thresholds; for missing values, filling them by upsampling method.
5. A prediction algorithm for predicting international crude oil prices according to claim 4, characterized in that: The upsampling method specifically comprises the following steps: S101: Divide the missing time period into n equal parts according to the minimum spacing of time points in the non-missing data time period, and obtain n-1 time points in the missing time period; S102: At n-1 time points, use the formula Get the data value P corresponding to the mth time point in the missing time period m , in formula (2), a0 is the data value corresponding to the last time point before the missing time period, a n is the data value corresponding to the first time point after the missing period.
6. A prediction algorithm for predicting international crude oil prices according to claim 5, characterized in that: In step S2, statistical software is used to fit and optimize the model to ensure the prediction accuracy and stability of the model.
7. A prediction algorithm for predicting international crude oil prices according to claim 6, characterized in that: The multiple regression equation in step S2 is: Y t = β0 + β1 × X1 + β2 × X2 + β3 × X3 + β4 × X4 + β5 × X5 + ε Equation (3) In formula (3), Y t The dependent variable is the international crude oil price in the tth period; β0 is the intercept term; β1, β2, β3, β4, and β5 are the regression coefficients of the influencing factors: US dollar exchange rate, global economic growth rate, oil production, geopolitical risk index, and supply and demand relationship, indicating the degree of impact on the international crude oil price when each influencing factor changes by one unit; ε is the error term, representing the part that cannot be explained in the model. ε is the difference between the actual result and the predicted result, which is included in the model to explain this subtle change.
8. A prediction algorithm for predicting international crude oil prices according to claim 7, characterized in that: Using the collected historical data, the historical data were standardized and converted, and statistical software was used to perform regression analysis on each influencing factor to obtain the regression model coefficients of β1, β2, β3, β4, and β5.
9. A prediction algorithm for predicting international crude oil prices according to claim 8, characterized in that: The statistical software includes using Python's statsmodels library or R language to perform regression analysis, and using regression analysis to calculate the estimated value of each regression coefficient.
10. The prediction algorithm for predicting international crude oil prices according to claim 9, characterized in that: The evaluation model in step S3 is built using open source Hadoop ecosystem components and Python programs.
