Copper future price rolling prediction method based on integrated machine learning
Through the method of rolling copper futures price prediction based on integrated machine learning, the problems of low price prediction accuracy and insufficient timeliness in the existing technology are solved, and the prediction effect of high precision and dynamic adaptation is achieved, and the company's market decision-making ability is improved.
Patent Information
- Application Number
- CN202510077684.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-01-17
- Publication Date
- 2025-05-16
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The prior art has problems such as low accuracy, insufficient timeliness and difficulty in dealing with market fluctuations in price predictions in the copper futures market.
The copper futures price rolling prediction method based on integrated machine learning is adopted to acquire and preprocess the input data, build a model intelligent matching network, generate high-precision prediction results, and combine the drawdown hedging model and event data verification to carry out investment planning.
It realizes high-precision copper futures price prediction, dynamically adapts to market changes, improves the decision-making ability of enterprises in rapidly changing markets, and ensures flexibility and forward-looking in dealing with market fluctuations.
Smart Images
Figure CN120013567A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of risk prediction, and in particular to a copper futures price rolling prediction method based on integrated machine learning. Background Art
[0002] Procurement forecasting refers to the process in which the decision makers of an enterprise, based on the data obtained from the investigation of the commodity procurement market, analyze and study, and use scientific methods to estimate the supply and demand of the commodity market and its changing trends in a certain period of time in the future, thereby providing a scientific basis for commodity procurement decisions and formulating commodity procurement plans, and achieving a series of goals such as sales profits.
[0003] In corporate procurement business, the existing procurement forecast has low price forecast accuracy, and the release time does not match the decision-making cycle, which affects the rationality of the annual planned price, especially when the market fluctuates greatly. The copper futures market is affected by complex factors such as dynamic nonlinearity, data noise, human manipulation and policy intervention, making it difficult to predict. Summary of the invention
[0004] Based on this, it is necessary to provide a copper futures price rolling forecasting method based on integrated machine learning to address the above technical problems.
[0005] A copper futures price rolling forecasting method based on integrated machine learning comprises the following steps:
[0006] Acquire input data, and obtain candidate feature variables according to the input data;
[0007] Build a model intelligent matching network;
[0008] Inputting the candidate feature variables into the model intelligent matching network, and obtaining a prediction result through the model intelligent matching network;
[0009] Develop a drawdown hedging model based on the forecast results;
[0010] Acquiring event data, verifying the validity of the event data, and obtaining adjusted event data;
[0011] An investment plan is obtained based on the adjustment event data and the drawdown hedging model for investment.
[0012] In one embodiment, obtaining input data, and obtaining candidate feature variables according to the input data includes:
[0013] Acquire input data, and preprocess the input data to obtain preprocessed data;
[0014] Expanding the data dimension of the preprocessed data by using a dimension-increasing tool to generate pre-candidate feature variables;
[0015] Obtain several prediction models;
[0016] By using a feature selection method, the pre-candidate feature variables are added or deleted based on the prediction model to obtain candidate feature variables.
[0017] In one embodiment, building a model intelligent matching network includes:
[0018] Testing the prediction accuracy of the prediction model, and in response to the prediction accuracy of the prediction model within a preset time range being greater than a predetermined prediction accuracy, using the prediction model as the prediction model in the model intelligent matching network;
[0019] In response to a prediction accuracy of the prediction model within a preset time range being less than a predetermined prediction accuracy, removing the prediction model from the model intelligent matching network;
[0020] The prediction accuracy of all prediction models in the model intelligent matching network is averaged.
[0021] In one embodiment, formulating a drawdown hedging model according to the prediction result includes:
[0022] Get preset hedging strategies;
[0023] Comparing the predicted result with the market price to obtain the probability of the predicted result exceeding the market price;
[0024] The hedging amount is calculated based on the prediction result exceedance probability and the preset hedging strategy.
[0025] In one embodiment, calculating the hedging amount according to the prediction result exceedance probability and the hedging strategy includes:
[0026] According to the predicted result exceeding probability, the differentiated hedging amount is calculated by the preset hedging strategy, and the differentiated hedging amounts are summed to obtain the hedging amount. In one embodiment, obtaining event data, verifying the validity of the event data, and obtaining the adjusted event data includes:
[0027] Using web crawler technology to obtain event data; wherein the event data includes: relevant policy events, financial news, social media comments and image data;
[0028] Obtain events that have a significant impact on the market from the event data, implement hedging verification, and obtain the potential market impact of the event data and its effectiveness in different scenarios;
[0029] In response to the validity of the event data being greater than a preset validity threshold, the event data is used as adjustment event data.
[0030] A copper futures price rolling forecasting system based on integrated machine learning, used to implement the copper futures price rolling forecasting method based on integrated machine learning as described above, comprising:
[0031] A feature acquisition module, used to acquire input data and obtain candidate feature variables according to the input data;
[0032] Matching construction module, used to build model intelligent matching network;
[0033] A prediction acquisition module, used for inputting the candidate feature variables into the model intelligent matching network, and obtaining a prediction result through the model intelligent matching network;
[0034] A drawdown formulation module, used to formulate a drawdown hedging model according to the prediction results;
[0035] A time acquisition module is used to acquire event data, verify the validity of the event data, and obtain adjusted event data;
[0036] The planning and prediction module is used to obtain an investment plan for investment based on the adjustment event data and the drawdown hedging model.
[0037] A device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of a method for rolling prediction of copper futures prices based on integrated machine learning described in each of the above embodiments are implemented.
[0038] A storage medium stores a computer program, which, when executed by a processor, implements the steps of a method for rolling prediction of copper futures prices based on integrated machine learning described in the above-mentioned embodiments.
[0039] Compared with the prior art, the advantages and beneficial effects of the present invention are: the present invention can efficiently handle the complexity of high-dimensional data and generate high-precision prediction results. Through daily rolling updates, it can dynamically adapt to market changes and achieve real-time predictions, thereby making up for the shortcomings of insufficient prediction accuracy and delayed timeliness of traditional corporate strategic systems and business newspaper publishing. While addressing the challenges of high-dimensional data, this method optimizes prediction accuracy, significantly improves the ability of enterprises to make strategic decisions in a rapidly changing market environment, and ensures their flexibility and foresight in responding to market fluctuations. This forecasting framework provides enterprises and market participants with more accurate and timely decision-making support tools. BRIEF DESCRIPTION OF THE DRAWINGS
[0040] Figure 1 A schematic diagram of a process of a copper futures price rolling forecasting method based on integrated machine learning in one embodiment;
[0041] Figure 2A schematic diagram of the output results of a CatBoost prediction model in one embodiment;
[0042] Figure 3 A schematic diagram of a hedging process in an embodiment;
[0043] Figure 4 A schematic diagram of the structure of a copper futures price rolling prediction system based on integrated machine learning in one embodiment;
[0044] Figure 5 Schematic diagram of the internal structure of a device in one embodiment. DETAILED DESCRIPTION
[0045] Before describing the specific embodiments of the present invention, the overall concept of the present invention is described as follows:
[0046] The present invention is mainly developed for the copper futures market forecasting process. The price forecasting accuracy of existing commercial newspapers is low, and the release time does not match the decision-making cycle, which affects the rationality of the annual planned price, especially when the market fluctuates greatly. The copper futures market is affected by complex factors such as dynamic nonlinearity, data noise, human manipulation and policy intervention, and the forecasting difficulty is high.
[0047] Therefore, the present invention proposes a copper futures price rolling forecasting method based on integrated machine learning, which processes high-dimensional data through integrated machine learning technology and combines it with a rolling forecasting model to achieve high-precision forecasting of copper futures prices.
[0048] After introducing the overall concept of the present invention, in order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below through specific implementation methods in conjunction with the accompanying drawings.
[0049] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in one or more embodiments of this specification should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in one or more embodiments of this specification do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.
[0050] In one embodiment, Figure 1 As shown, a copper futures price rolling forecasting method based on integrated machine learning is provided, comprising the following steps:
[0051] Step S101, acquiring input data, and obtaining candidate feature variables according to the input data.
[0052] Specifically, by integrating global financial markets, macroeconomic indicators and copper supply and demand data, a multidimensional data framework was constructed, and real-time data access was achieved using API interfaces to ensure the comprehensiveness and timeliness of input data. The input data was processed to obtain candidate feature variables.
[0053] On this basis, input data is obtained, and candidate feature variables are obtained according to the input data, including:
[0054] Acquire input data, and preprocess the input data to obtain preprocessed data;
[0055] Expanding the data dimension of the preprocessed data by using a dimension-increasing tool to generate pre-candidate feature variables;
[0056] Obtain several prediction models;
[0057] By using a feature selection method, the pre-candidate feature variables are added or deleted based on the prediction model to obtain candidate feature variables.
[0058] Specifically, a standardized preprocessing process is used for input data to solve the time inconsistency and missing problems, and the time granularity is adjusted according to the characteristics of the input data. Dimensional feature engineering is used to expand the input data dimension and generate pre-candidate feature variables.
[0059] Obtain several prediction models and use the grid search method to tune the hyperparameters of each prediction model. Among them, the prediction models include: NGBoost, LightGBM, XGBoost, CatBoost, LSTM, CNN, RNN and more than 20 other models.
[0060] For the pre-candidate feature variables, the feature selection method is used to add or delete pre-candidate feature variables based on the prediction model, and the pre-candidate feature variables are screened from the alternative pre-candidate feature variables, and iterated multiple times. Finally, the core pre-candidate feature variables with high contribution to the model performance are retained, and the pre-candidate feature variables with low contribution or redundant are removed. The output of the CatBoost prediction model is as follows Figure 2 The top ten candidate feature variables of CatBoost importance are shown in Table 1 (due to the large number of models and candidate feature variables involved, only some are selected for display).
[0061] Table 1 CatBoost top ten candidate feature variables
[0062] Data Name Importance Shanghai Composite Index 20.18475157 Shanghai Composite Index 12.16284108 OECD Leading Economic Index(China) 4.583216449 Consumer Confidence Index (US) 4.336299667 US long-short interest rate differential 3.723027765 Rio Tinto 3.388399954 Australia 10-year bond yield 3.234400784 Purchasing Managers Index (China) 3.032517046 Crude steel output (output) 2.994303908 OECD Leading Economic Index (Australia) 2.897752709
[0063] Step S102, constructing a model intelligent matching network.
[0064] Specifically, more than 20 prediction model frameworks such as NGBoost, LightGBM, XGBoost, CatBoost, LSTM, CNN, and RNN are included in the network candidates. The prediction model is selected based on the prediction results, and a model intelligent matching network is constructed.
[0065] On this basis, building a model intelligent matching network includes:
[0066] Testing the prediction accuracy of the prediction model, and in response to the prediction accuracy of the prediction model within a preset time range being greater than a predetermined prediction accuracy, using the prediction model as the prediction model in the model intelligent matching network;
[0067] In response to a prediction accuracy of the prediction model within a preset time range being less than a predetermined prediction accuracy, removing the prediction model from the model intelligent matching network;
[0068] The prediction accuracy of all prediction models in the model intelligent matching network is averaged.
[0069] Specifically, the candidate feature variables are input into the model for model training. The training results are output and the results of different prediction models are compared. The prediction models with a prediction accuracy of more than 65% are added to the model intelligent matching network. According to the scheduled time, after multiple iterations and optimizations, until the accuracy of all prediction models in the model intelligent matching network reaches the target level (65%), the prediction results are averaged to obtain a model intelligent matching network that can be used for hedging analysis.
[0070] In this embodiment, it is finally shown that the best performing prediction model is CatBoost, which can achieve a prediction accuracy of more than 65%. Therefore, CatBoost is added to the model intelligent matching network, and the prediction performance comparison is shown in Table 2 (due to the involvement of many models, only some are selected for display).
[0071] Table 2 Comparison of prediction performance
[0072] Prediction Set CatBoost LightGBM NGBoost XGBoost Rise and fall accuracy 0.765 0.623 0.621 0.640 R squared 0.420 0.084 0.308 -1.561
[0073] Step S103, inputting the candidate feature variables into the model intelligent matching network, and obtaining a prediction result through the model intelligent matching network.
[0074] Specifically, the candidate feature variables are input into the model intelligent matching network to obtain the prediction results.
[0075] Step S104, selecting a drawdown hedging model according to the prediction result.
[0076] Specifically, a preset drawdown hedging model is obtained, and the drawdown hedging model is selected according to the prediction result.
[0077] On this basis, the drawdown hedging model is formulated according to the forecast results, including:
[0078] Get preset hedging strategies;
[0079] Comparing the predicted result with the market price to obtain the probability of the predicted result exceeding the market price;
[0080] The hedging amount is calculated based on the prediction result exceedance probability and the preset hedging strategy.
[0081] On this basis, the hedging amount is calculated according to the prediction result exceeding probability and the hedging mode, including:
[0082] According to the probability of exceeding the predicted result, the differentiated hedging amounts are calculated through the preset hedging strategy, and the differentiated hedging amounts are summed up to obtain the hedging amount.
[0083] Specifically, the following preset hedging strategies are defined based on enterprise needs: Prerequisite: The hedging range is within 100% of the required purchase volume, and two preset hedging strategies are used at the same time.
[0084] Strategy 1: The goal is to hedge against changes in purchase prices and reduce purchase costs, according to the traditional hedging perspective.
[0085] Step 1: Determine the trading opportunity: anytime.
[0086] Step 2: Hedging judgment conditions: judge based on the probability that the predicted value of the physical price exceeds the predicted result of the futures price. The discriminant formula refers to Table 3.
[0087] Step 3: Determine the transaction volume, which is determined by the probability that the market price exceeds the predicted value, and the maximum transaction volume is 40%.
[0088] Table 3 Hedging strategy-discriminant reference table
[0089]
[0090] Among them, F represents the 6-month futures price, S represents the current physical price, exp() represents the exponential function, μ represents the predicted value of the logarithmic rate of change of the physical price after 6 months, and σ represents the predicted value of the volatility of the physical price after 6 months.
[0091] Model 2: Hedge the stable purchase of raw materials with reference to the forecast price, based on the risk management perspective.
[0092] Step 1: Determine the timing of the transaction: once a month or when the same situation lasts for a week, etc.
[0093] Step 2: Hedging judgment condition: judge based on the probability that the predicted value of the entity price exceeds the predicted price. The discriminant formula refers to Table 4.
[0094] Step 3: Determine the transaction volume: The transaction volume is determined based on the probability that the market price exceeds the predicted value. The maximum transaction volume is 40%.
[0095] Table 4 Reference table of hedging strategy two discriminants
[0096]
[0097] Among them, F represents the 6-month futures price, T represents the predicted price, S represents the current physical price, Min() represents the minimum value, exp() represents the exponential function, μ represents the predicted value of the logarithmic rate of change of the physical price after 6 months, and σ represents the predicted value of the volatility of the physical price after 6 months.
[0098] The preset hedging strategies are calculated and differentiated by hedging amount, and the differentiated hedging amounts of all preset hedging strategies are summed up to obtain the hedging amount.
[0099] The forecast results show the hedging flow chart of the drawdown hedging model as follows Figure 3 shown.
[0100] Step S105, acquiring event data, performing validity verification on the event data, and obtaining adjusted event data.
[0101] Specifically, we obtain event data, conduct in-depth analysis on events that have had a significant impact on the market in the past, implement hedge verification, and obtain adjusted event data.
[0102] On this basis, the event data is obtained, and the validity of the event data is verified, and the adjusted event data includes:
[0103] Using web crawler technology to obtain event data; wherein the event data includes: relevant policy events, financial news, social media comments and image data;
[0104] Obtain events that have a significant impact on the market from the event data, implement hedging verification, and obtain the potential market impact of the event data and its effectiveness in different scenarios;
[0105] In response to the validity of the event data being greater than a preset validity threshold, the event data is used as adjustment event data.
[0106] Specifically, we use web crawler technology to capture relevant policy events, financial news, social media comments and image data, and process the acquired data, including data cleaning and labeling (using tensorflow for labeling), to build an event dataset.
[0107] Obtain events that have a significant impact on the market from the event data, implement hedging verification, and obtain the potential market impact of the event data and its effectiveness in different scenarios. When the effectiveness of the event data is greater than the preset effectiveness threshold, the event data will be used as adjusted event data.
[0108] Step S106, obtaining an investment plan and investing based on the adjustment event data and the drawdown hedging model.
[0109] Specifically, we invest based on adjustment event data and drawdown hedging models to optimize the risk management and return potential of the portfolio. The data is updated on a rolling basis at 9 a.m. every day, and can be automatically updated according to specific needs to ensure that the latest market dynamics are always reflected.
[0110] The present application provides a copper futures price rolling forecasting method based on integrated machine learning, which establishes an effective feature screening mechanism, adds or deletes the input data of the model intelligent matching network, improves the accuracy, and retains candidate feature variables that contribute more to the prediction, such as the Shanghai Composite Index, the Shanghai Composite Index, the OECD Leading Economic Index (China), etc., while removing candidate feature variables that contribute very little to the prediction performance, such as the number of new infections in various countries, etc., which can significantly improve the prediction accuracy of the model intelligent matching network.
[0111] It provides a suitable model selection method, incorporating more than 20 prediction models such as NGBoost, LightGBM, XGBoost, CatBoost, LSTM, CNN, and RNN into the model intelligent matching network candidates, and then simultaneously evaluates the prediction model performance to select the most suitable prediction model.
[0112] A scientific hedging strategy has been formulated. Strategy 1: Based on the distribution of copper prices in 6 months, in response to the possibility that the physical price may exceed the futures price, a strategy of purchasing raw materials at a low price is adopted. The lower the futures price, the greater the amount of hedging. Strategy 2: Based on the distribution of copper prices in 6 months, in order to respond to the forecast results (forecasted prices) or the possibility of futures prices exceeding, a hedging strategy is adopted. The higher the possibility (probability) of the forecast price exceeding, the higher the priority of raw material procurement, and the greater the amount of hedging should be. Based on this strategy, the risk of future copper price fluctuations can be effectively reduced.
[0113] The present invention provides a copper futures price rolling forecasting method based on an integrated algorithm, which can efficiently handle the complexity of high-dimensional data and generate high-precision forecasting results. Through daily rolling updates, it can dynamically adapt to market changes and achieve real-time forecasting, thereby making up for the shortcomings of insufficient forecasting accuracy and delayed timeliness of traditional corporate strategic systems and business newspaper publishing. While responding to the challenges of high-dimensional data, this method optimizes forecasting accuracy, significantly improves the ability of enterprises to make strategic decisions in a rapidly changing market environment, and ensures their flexibility and foresight in responding to market fluctuations. This forecasting framework provides enterprises and market participants with more accurate and timely decision-making support tools.
[0114] It should be noted that the method of the embodiment of the present invention can be performed by a single device, such as a computer or a server. The method of this embodiment can also be applied in a distributed scenario and completed by multiple devices cooperating with each other. In the case of such a distributed scenario, one of the multiple devices can only perform one or more steps in the method of the embodiment of the present invention, and the multiple devices will interact with each other to complete the described method.
[0115] It should be noted that some embodiments of the present invention are described above. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the above embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0116] Based on the same inventive concept, corresponding to any of the above-mentioned embodiment methods, the present invention also provides a copper futures price rolling prediction system based on integrated machine learning.
[0117] refer to Figure 4 , the copper futures price rolling forecasting system based on integrated machine learning comprises:
[0118] A feature acquisition module 401 is used to acquire input data and obtain candidate feature variables according to the input data;
[0119] A matching construction module 402 is used to construct a model intelligent matching network;
[0120] A prediction acquisition module 403 is used to input the candidate feature variables into the model intelligent matching network and obtain a prediction result through the model intelligent matching network;
[0121] A drawdown formulation module 404, used to formulate a drawdown hedging model according to the prediction results;
[0122] The time acquisition module 405 is used to acquire event data, verify the validity of the event data, and obtain adjustment event data;
[0123] The planning and prediction module 406 is used to obtain an investment plan for investment based on the adjustment event data and the drawdown hedging model.
[0124] For the convenience of description, the above system is described as being divided into various modules according to their functions. Of course, when implementing the present invention, the functions of each module can be implemented in the same or multiple software and / or hardware.
[0125] The system of the above embodiment is used to implement a corresponding copper futures price rolling prediction method based on integrated machine learning in any of the above embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.
[0126] Based on the same inventive concept, corresponding to any of the above-mentioned embodiments and methods, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, a rolling prediction method for copper futures prices based on integrated machine learning as described in any of the above embodiments is implemented.
[0127] Figure 5 A more specific schematic diagram of the hardware structure of an electronic device provided in this embodiment is shown, and the device may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040 are connected to each other through the bus 1050 in the device.
[0128] The processor 1010 can be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.
[0129] The memory 1020 may be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 may store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program codes are stored in the memory 1020 and are called and executed by the processor 1010.
[0130] The input / output interface 1030 is used to connect the input / output module to realize information input and output. The input / output module can be configured in the device as a component (not shown in the figure), or it can be externally connected to the device to provide corresponding functions. The input device may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output device may include a display, a speaker, a vibrator, an indicator light, etc.
[0131] The communication interface 1040 is used to connect a communication module (not shown) to realize communication interaction between the device and other devices. The communication module can realize communication through a wired mode (such as USB, network cable, etc.) or a wireless mode (such as mobile network, WIFI, Bluetooth, etc.).
[0132] The bus 1050 includes a path that transmits information between the various components of the device (eg, the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040).
[0133] It should be noted that, although the above device only shows the processor 1010, the memory 1020, the input / output interface 1030, the communication interface 1040 and the bus 1050, in the specific implementation process, the device may also include other components necessary for normal operation. In addition, it can be understood by those skilled in the art that the above device may also only include the components necessary for implementing the embodiments of the present specification, and does not necessarily include all the components shown in the figure.
[0134] The electronic device of the above embodiment is used to implement a corresponding copper futures price rolling prediction method based on integrated machine learning in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be repeated here.
[0135] Based on the same inventive concept, corresponding to any of the above-mentioned embodiment methods, the present invention also provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute a copper futures price rolling forecasting method based on integrated machine learning as described in any of the above embodiments.
[0136] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device.
[0137] The computer instructions stored in the storage medium of the above embodiment are used to enable the computer to execute a copper futures price rolling forecasting method based on integrated machine learning as described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.
[0138] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present invention (including the claims) is limited to these examples. Under the concept of the present invention, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the different aspects of the embodiments of the present invention as described above, which are not provided in detail for the sake of simplicity.
[0139] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present invention belong.
[0140] In addition, to simplify the description and discussion, and in order not to obscure the embodiments of the present invention, known power / ground connections to integrated circuit (IC) chips and other components may or may not be shown in the provided figures. In addition, the devices may be shown in the form of block diagrams to avoid obscuring the embodiments of the present invention, and this also takes into account the fact that the details of the implementation of these block diagram devices are highly dependent on the platform on which the embodiments of the present invention will be implemented (i.e., these details should be fully within the scope of understanding of those skilled in the art). Where specific details (e.g., circuits) are set forth to describe exemplary embodiments of the present invention, it will be apparent to those skilled in the art that embodiments of the present invention may be implemented without these specific details or with variations in these specific details. Therefore, these descriptions should be considered illustrative rather than restrictive.
[0141] Although the invention has been described in conjunction with specific embodiments of the invention, many replacements, modifications and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may use the embodiments discussed.
[0142] The embodiments of the present invention are intended to cover all such substitutions, modifications and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present invention should be included in the protection scope of the present invention.
Claims
1. A copper futures price rolling forecasting method based on integrated machine learning, characterized in that: include: Acquire input data, and obtain candidate feature variables according to the input data; Build a model intelligent matching network; Inputting the candidate feature variables into the model intelligent matching network, and obtaining a prediction result through the model intelligent matching network; Develop a drawdown hedging model based on the forecast results; Acquiring event data, verifying the validity of the event data, and obtaining adjusted event data; An investment plan is obtained based on the adjustment event data and the drawdown hedging model for investment.
2. A copper futures price rolling forecasting method based on integrated machine learning according to claim 1, characterized in that: The acquiring of input data and obtaining candidate feature variables according to the input data comprises: Acquire input data, and preprocess the input data to obtain preprocessed data; Expanding the data dimension of the preprocessed data by using a dimension-increasing tool to generate pre-candidate feature variables; Obtain several prediction models; By using a feature selection method, the pre-candidate feature variables are added or deleted based on the prediction model to obtain candidate feature variables.
3. A copper futures price rolling forecasting method based on integrated machine learning according to claim 2, characterized in that: The construction of the model intelligent matching network includes: Testing the prediction accuracy of the prediction model, and in response to the prediction accuracy of the prediction model within a preset time range being greater than a predetermined prediction accuracy, using the prediction model as the prediction model in the model intelligent matching network; In response to a prediction accuracy of the prediction model within a preset time range being less than a predetermined prediction accuracy, removing the prediction model from the model intelligent matching network; The prediction accuracy of all prediction models in the model intelligent matching network is averaged.
4. A copper futures price rolling forecasting method based on integrated machine learning according to claim 1, characterized in that: The formulating of the drawdown hedging model according to the prediction results comprises: Get preset hedging strategies; Comparing the predicted result with the market price to obtain the probability of the predicted result exceeding the market price; The hedging amount is calculated based on the prediction result exceedance probability and the preset hedging strategy.
5. A copper futures price rolling forecasting method based on integrated machine learning according to claim 4, characterized in that: The calculating of the hedging amount according to the prediction result exceeding probability and the hedging strategy comprises: According to the probability of exceeding the predicted result, the differentiated hedging amounts are calculated through the preset hedging strategy, and the differentiated hedging amounts are summed up to obtain the hedging amount.
6. A copper futures price rolling forecasting method based on integrated machine learning according to claim 1, characterized in that: The acquiring event data and verifying the validity of the event data to obtain the adjusted event data includes: Using web crawler technology to obtain event data; wherein the event data includes: relevant policy events, financial news, social media comments and image data; Obtain events that have a significant impact on the market from the event data, implement hedging verification, and obtain the potential market impact of the event data and its effectiveness in different scenarios; In response to the validity of the event data being greater than a preset validity threshold, the event data is used as adjustment event data.
7. A copper futures price rolling forecasting system based on integrated machine learning, characterized in that: A method for rolling forecasting of copper futures prices based on integrated machine learning as described in any one of claims 1 to 6, comprising: A feature acquisition module, used to acquire input data and obtain candidate feature variables according to the input data; Matching construction module, used to build model intelligent matching network; A prediction acquisition module, used for inputting the candidate feature variables into the model intelligent matching network, and obtaining a prediction result through the model intelligent matching network; A drawdown formulation module, used to formulate a drawdown hedging model according to the prediction results; A time acquisition module is used to acquire event data, verify the validity of the event data, and obtain adjusted event data; The planning and prediction module is used to obtain an investment plan for investment based on the adjustment event data and the drawdown hedging model.
8. A device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.
9. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.