An intelligent financial engineering quantitative research and reproduction method and related components

Abstract

The present invention discloses an intelligent financial engineering quantitative research and reproduction method and related components, which relate to the financial field. First, the target workflow is determined by combining the workflow template and the metadata extracted from the quantitative research document by the information extraction agent; when executing the target workflow, the information extraction agent extracts the quantitative calculation information from the document; the code agent generates the quantitative calculation code according to the quantitative calculation information and the data source information, and obtains the quantitative factor data after executing the quantitative calculation code; the quantitative factor data is quantitatively backtested according to the backtesting tool corresponding to the backtesting method specified in the document to obtain the backtesting results; the verification agent compares the backtesting results with the quantitative research results in the document to draw the quantitative research conclusions. The present invention automatically completes the entire process of financial engineering quantitative research based on the large model agent, allowing practitioners to focus on the core work of quantitative design and strategy design, and improve the research efficiency of users.

Description

Technical Field

[0001] The present invention relates to the financial field, and in particular to an intelligent financial engineering quantitative research and reproduction method and related components. Background Art

[0002] Quantitative financial engineering research utilizes mathematics, statistics, and computer technology to analyze and process various financial market data, building quantitative models to formulate investment strategies and manage risk. Within the financial industry, quantitative financial engineering research is a relatively high-level field, requiring not only a deep understanding of financial quantitative expertise but also a familiarity with supporting infrastructure (such as specialized market data and computing services). This places high demands on practitioners. While designing, implementing, and validating their own quantitative financial research projects, practitioners also need to stay abreast of market and research trends, reading and replicating numerous existing research projects and reports. This workload is extensive and requires a high level of professional expertise.

[0003] It can be seen from this that even for industry insiders, the industry threshold for conducting financial engineering quantitative research is relatively high. Therefore, there is an urgent need for a financial engineering quantitative research method that can improve work efficiency and reduce research difficulty. Summary of the Invention

[0004] The purpose of this invention is to provide an intelligent financial engineering quantitative research and reproduction method and related components, which can complete the entire process of financial engineering quantitative research based on a large model intelligent body, allowing practitioners to focus on the core work of quantitative design and strategy design, and improve users' research efficiency.

[0005] To solve the above technical problems, the present invention provides an intelligent financial engineering quantitative research and replication method, which includes:

[0006] Utilize an information extraction agent to extract meta-information related to the workflow structure from the quantitative research document provided by the user, and determine and execute the target workflow by combining the meta-information with the workflow template;

[0007] When executing the target workflow, the quantitative research document is input into the information extraction agent, and the quantitative calculation information extracted from the quantitative research document by the information extraction agent is obtained;

[0008] Inputting the quantitative calculation information into a code agent so that the code agent generates quantitative calculation code based on the quantitative calculation information and data source information obtained by calling a professional domain knowledge base, a data warehouse, and professional calculation tools through an interface agent, obtains professional data through the interface agent, and executes the quantitative calculation code to generate quantitative factor data;

[0009] Extracting the backtesting method specified in the quantitative research document according to the information extraction agent, calling the backtesting tool corresponding to the backtesting method through the interface agent, and performing quantitative backtesting on the quantitative factor data using the backtesting tool to obtain a backtesting result;

[0010] The backtest results and the quantitative research results extracted by the information extraction agent from the quantitative research document are input into the verification agent to obtain the quantitative research conclusions drawn by the verification agent by comparing the backtest results with the quantitative research results.

[0011] Optionally, the information extraction agent extracts quantitative calculation information from the quantitative research document, including:

[0012] The information extraction agent sets prompt words related to the quantitative calculation information in a step-by-step manner from macro to detail, and extracts the quantitative calculation information from the quantitative research document according to the prompt words in order from macro to detail;

[0013] And / or, the information extraction agent sets prompt words related to the quantitative calculation information from a plurality of preset dimensions, and extracts the quantitative calculation information from the quantitative research document according to each of the prompt words.

[0014] Optionally, the information extraction agent extracts quantitative calculation information from the quantitative research document according to each of the prompt words, including:

[0015] The information extraction agent uses two or more different types of large models to extract quantitative calculation information from the quantitative research document respectively;

[0016] The quantitative calculation information extracted by each of the large models from the quantitative research document is deduplicated and integrated to obtain the final quantitative calculation information, and when the difference between the quantitative calculation information extracted by each large model reaches a preset threshold, a prompt message is generated to prompt the user to correct the quantitative calculation information.

[0017] Optionally, after the information extraction agent extracts the quantitative calculation information from the quantitative research document according to each of the prompt words, the information extraction agent further includes:

[0018] Determining whether the currently extracted quantitative calculation information satisfies a preset completeness verification condition, where the preset completeness verification condition is a condition corresponding to when the quantitative calculation information satisfies information required by the code agent to generate the quantitative calculation code;

[0019] If so, the step of inputting the quantitative calculation information into the code agent is performed so that the code agent generates quantitative calculation code based on the quantitative calculation information and the data source information obtained by calling the professional domain knowledge base, data warehouse and professional calculation tools through the interface agent;

[0020] If not, the process proceeds to the step of extracting the quantitative calculation information from the quantitative research document according to each of the prompt words.

[0021] Optionally, the code agent generates quantitative calculation code based on the quantitative calculation information and data source information obtained by the interface agent by calling a professional domain knowledge base, a data warehouse, and a professional calculation tool, including:

[0022] The code agent generates a quantitative calculation code according to the quantitative calculation information and a preset prompt word, and calls a professional domain knowledge base through the interface agent when generating the quantitative calculation code to adjust the quantitative calculation code in combination with the content of the professional domain knowledge base;

[0023] The code agent calls the data warehouse through the interface agent to obtain relevant data required for sample testing of the quantitative calculation code, and calls the professional calculation tool through the interface agent to test the quantitative calculation code using the test sample;

[0024] When the quantitative calculation code passes the sample test, the quantitative calculation code is stored in a library as the final quantitative calculation code;

[0025] When the quantitative calculation code fails the sample test, the error and exception information that occurs during the execution of the quantitative calculation code will be compiled into the professional field knowledge base, and the process of generating the quantitative calculation code according to the preset prompt words based on the quantitative calculation information will be re-entered, and when generating the quantitative calculation code, the professional field knowledge base will be called through the interface intelligent body to adjust the quantitative calculation code in combination with the content in the professional field knowledge base.

[0026] Optionally, after the code agent executes the quantization calculation code to generate quantization factor data, the method further includes:

[0027] The information extraction agent extracts the backtesting method specified in the quantitative research document, and calls the backtesting tool corresponding to the backtesting method through the interface agent. The backtesting tool is used to perform quantitative testing on the quantitative factor data in combination with the current real market data to obtain the current test results, so as to determine the performance of the quantitative factor data in the current market.

[0028] Optionally, the interface agent can call professional domain knowledge bases, data warehouses, professional calculation tools, and backtesting tools, including:

[0029] Taking each first candidate tool in the first layer of the target access object as a current candidate tool, and writing each current candidate tool into a function call parameter of the interface agent so that the interface agent selects the current target tool from each current candidate tool;

[0030] The target access objects include the professional domain knowledge base, the data warehouse, the professional computing tools, and the backtesting tools, and the professional domain knowledge base, the data warehouse, the professional computing tools, and the backtesting tools are all reconstructed into a tree structure in advance according to a preset hierarchical standard;

[0031] The alternative tools in the next layer of the node where the current target tool is located are used as new current alternative tools, and the step of writing each current alternative tool into the function call parameter of the interface agent is entered so that the interface agent can select the current target tool from the current alternative tools. Until the current target tool is the actual tool, the structured parameters required to call the actual tool are returned to the interface agent so that the interface agent can call the target access object to complete the corresponding function.

[0032] To solve the above technical problems, the present invention also provides an intelligent financial engineering quantitative research and reproduction system, which includes:

[0033] a target workflow determination unit, configured to extract meta-information related to the workflow structure from the quantitative research document provided by the user using an information extraction agent, and determine and execute the target workflow by combining the meta-information with the workflow template;

[0034] a quantitative calculation information extraction unit, configured to input the quantitative research document into the information extraction agent when executing the target workflow, and obtain the quantitative calculation information extracted from the quantitative research document by the information extraction agent;

[0035] a quantization factor data generation unit, configured to input the quantization calculation information into a code agent, so that the code agent generates quantization calculation code based on the quantization calculation information and data source information obtained by calling a professional domain knowledge base, a data warehouse, and professional calculation tools through an interface agent, obtain professional data through the interface agent, and execute the quantization calculation code to generate quantization factor data;

[0036] A backtesting unit, configured to extract the backtesting method specified in the quantitative research document based on the information extraction agent, call the backtesting tool corresponding to the backtesting method through the interface agent, and use the backtesting tool to perform quantitative backtesting on the quantitative factor data to obtain a backtesting result;

[0037] A verification unit is used to input the backtest results and the quantitative research results extracted by the information extraction agent from the quantitative research document into a verification agent, and obtain a quantitative research conclusion drawn by the verification agent by comparing the backtest results with the quantitative research results.

[0038] To solve the above technical problems, the present invention also provides an intelligent financial engineering quantitative research and reproduction device, which includes:

[0039] memory for storing computer programs;

[0040] A processor is configured to implement the steps of any of the above-mentioned intelligent financial engineering quantitative research and reproduction methods when executing the computer program.

[0041] In order to solve the above technical problems, the present invention also provides a storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of any of the above-mentioned intelligent financial engineering quantitative research and reproduction methods are implemented.

[0042] The beneficial effect of the present invention is to provide an intelligent financial engineering quantitative research and reproduction method and related components, systems, devices and storage media. First, the target workflow is determined by combining the workflow template and the metadata extracted from the quantitative research document by the information extraction agent. When executing the target workflow, the information extraction agent extracts the quantitative calculation information from the document; the code agent generates the quantitative calculation code based on the quantitative calculation information and the data source information, and obtains the quantitative factor data after executing the quantitative calculation code; the quantitative factor data is quantitatively backtested according to the backtesting tool corresponding to the backtesting method specified in the document to obtain the backtesting results; the verification agent compares the backtesting results with the research results in the document to draw the quantitative research conclusions. The entire process of financial engineering quantitative research is completed based on the large model agent, allowing practitioners to focus on the core work of quantitative design and strategy design, thereby improving the research efficiency of users. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the prior art and the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0044] Figure 1 A first flow chart of an intelligent financial engineering quantitative research and replication method provided by the present invention;

[0045] Figure 2 A second flow chart of an intelligent financial engineering quantitative research and replication method provided by the present invention;

[0046] Figure 3 A workflow diagram of an information extraction agent provided by the present invention;

[0047] Figure 4 A workflow diagram of a code agent provided by the present invention;

[0048] Figure 5 A workflow diagram of an interface agent provided by the present invention;

[0049] Figure 6 A schematic diagram of the structure of an intelligent financial engineering quantitative research and reproduction system provided by the present invention;

[0050] Figure 7 This is a structural diagram of an intelligent financial engineering quantitative research and reproduction device provided by the present invention. DETAILED DESCRIPTION

[0051] The core of this invention is to provide an intelligent financial engineering quantitative research and reproduction method and related components, which can complete the entire process of financial engineering quantitative research based on a large model intelligent body, allowing practitioners to focus on the core work of quantitative design and strategy design, and improve users' research efficiency.

[0052] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0053] Please refer to Figure 1 , Figure 1 This is a first flow chart of an intelligent financial engineering quantitative research and replication method provided by the present invention, which includes:

[0054] S101. Utilize an information extraction agent to extract meta-information related to the workflow structure from the quantitative research document provided by the user, and determine and execute the target workflow by combining the meta-information and the workflow template.

[0055] First, it should be noted that the quantitative research document in the present invention can be a financial engineering quantitative research description document or a financial engineering research report on the market. A financial engineering quantitative research description document typically includes a detailed description of the quantitative research background, objectives, methodology, data sources, model building process, strategy implementation methods, and their application. Financial engineering research reports typically include market perspectives, strategic recommendations, and investment advice. Although financial engineering quantitative research description documents and financial engineering research reports differ in their focus, both include the data, methods, and conclusions required for conducting financial engineering quantitative research. The intelligent financial engineering quantitative research and replication method provided by the present invention can automatically complete financial engineering quantitative research based on the above content. In addition, it is recommended that users provide quantitative research documents in PDF file format.

[0056] Considering the diverse nature of quantitative research in financial engineering, such as stock selection, timing, derivatives, and fundamentals, each type of quantitative research requires different key information, such as source data schemas and testing methods. Consequently, the workflow details required to complete each type of quantitative research also differ. Therefore, it is necessary to first determine the target workflow for this quantitative research based on the user-provided quantitative research document.

[0057] Specifically, an information extraction agent is used to extract metadata related to the workflow structure from the quantitative research document. Meta-information related to the workflow structure in this invention includes, but is not limited to, the overall judgment of the quantitative research, the research type, source data information, and testing methods. The dynamic prompt of the information extraction agent must meet the aforementioned metadata requirements and conform to the JSON structured output format.

[0058] The information extraction agent is an intelligent system that uses artificial intelligence technology to automatically identify and extract specified information from quantitative research documents. It is capable of interacting with users in natural language, both in conversational and document-based formats. In this step, after inputting the user-provided quantitative research document into the information extraction agent, a prompt such as "Please extract metadata related to the workflow structure in this quantitative research document" can be input into the information extraction agent. The information extraction agent will then automatically extract the aforementioned metadata from the quantitative research document. There are many different types of information extraction agents capable of performing information extraction, and the present invention does not restrict their specific types; the agent can be selected based on actual needs.

[0059] Please refer to Figure 2 , Figure 2This is the second flow chart of the intelligent financial engineering quantitative research and replication method provided by the present invention. After a user-provided research document or report is input into the information extraction agent, the information extraction agent extracts metadata from the document (i.e., the user-provided quantitative research document, hereinafter referred to as the document). The present invention further utilizes a pre-configured workflow template to identify a workflow whose workflow structure conforms to the extracted metadata related to the workflow structure, and then identifies this workflow as the target workflow for this financial engineering quantitative research.

[0060] The workflow defines the steps required to automatically complete quantitative financial engineering research. Its main structure includes extracting quantitative calculation information using an information extraction agent, generating quantitative calculation code and quantitative factor data using a code agent, and performing quantitative testing and verification using specialized tools and verification agents. The following describes the process of completing quantitative research using different large-scale model agents according to the target workflow.

[0061] S102. When executing the target workflow, the quantitative research document is input into the information extraction agent to obtain the quantitative calculation information extracted from the quantitative research document by the information extraction agent.

[0062] First, an information extraction agent is needed to extract quantitative calculation information from quantitative research documents so that quantitative calculation code can be generated based on this information. Quantitative calculation information includes, but is not limited to, a description of the quantitative factor calculation logic, calculation formulas, step-by-step calculation autonomous planning, data source information, and factor structured metadata (frequency, etc.).

[0063] The process of an information extraction agent extracting quantitative computing information from a document is similar to the process of an information extraction agent extracting meta-information from a document, except that the prompt words are different. The present invention does not limit the specific implementation method of the information extraction agent extracting quantitative computing information from a document based on the prompt words.

[0064] However, it is important to emphasize that the present invention provides different prompt words from the macro to the detailed, as well as a series of prompt words related to quantitative calculation information from different angles. This enables the information extraction agent to extract quantitative calculation information from quantitative research documents in a step-by-step and multi-angle redundant extraction manner, ensuring the completeness of the quantitative calculation information and, in turn, ensuring that rigorous and professional quantitative calculation code can be generated based on the quantitative calculation information. The specific process will be explained in subsequent examples and will not be repeated here.

[0065] S103. Input the quantitative calculation information into the code agent so that the code agent generates quantitative calculation code based on the quantitative calculation information and the data source information obtained by calling the professional domain knowledge base, data warehouse and professional calculation tools through the interface agent, obtains professional data through the interface agent and executes the quantitative calculation code to generate quantitative factor data.

[0066] Generating quantitative calculation code based on quantitative calculation information is a core component of quantitative financial engineering research. Previously, manually writing quantitative calculation code was difficult, time-consuming, and labor-intensive. This invention utilizes a code agent to generate quantitative calculation code after obtaining quantitative calculation information. To ensure that the code agent can smoothly and accurately generate quantitative calculation code that meets the requirements, an interface agent is used to access specialized domain knowledge bases, data warehouses, and specialized calculation tools during the process of generating quantitative calculation code based on quantitative calculation information.

[0067] Specifically, the present invention pre-collects general information from the field of financial quantification based on RAG (Retrieval-Augmented Generation) and stores it in a professional domain knowledge base. When the code agent generates quantitative calculation code, it uses the interface agent to preferentially query the professional domain knowledge base for relevant professional information, thereby generating quantitative calculation code that meets industry requirements. Furthermore, several case studies of code errors that are prone to occur in financial data processing and quantitative calculation scenarios are collected and stored in the professional domain knowledge base. These cases serve as supplementary information to assist the code agent in generating correct and effectively executable quantitative calculation code.

[0068] In addition, a self-feedback iteration mechanism can be used to allow the code agent to self-diagnose and repair the current problem code. After multiple rounds of iterations, the problems of the quantitative calculation code generated in the previous round can be gradually repaired until the quantitative calculation code no longer reports errors and the results meet expectations.

[0069] The present invention further utilizes an interface agent to access data from a data warehouse and specialized computing tools, and uses test samples to test the generated quantitative computing code. If the quantitative computing code passes the sample test, it is considered to have passed the verification and is saved. If the quantitative computing code fails the sample test, the quantitative computing code is adjusted based on the problems encountered, and the sample test is re-performed on the adjusted quantitative computing code until the code passes the sample test. The quantitative computing code that passes the sample test is saved as the final quantitative computing code.

[0070] In summary, the code agent leverages the interface agent to access domain-specific knowledge bases, data warehouses, and specialized calculation tools, ensuring that the quantitative calculation code generated based on quantitative calculation information meets requirements such as the description of quantitative factor calculation logic, alignment of data and variable field names, and timely invocation of industry-specific data and specialized calculation tools. The data source schema information determined by the dynamic workflow template will dynamically update the prompt information here, ensuring that the generated quantitative calculation code reflects core information such as the correct data dictionary. In conjunction with the domain-specific knowledge base, specialized descriptions of algorithms such as financial concepts and indicators, which are prone to ambiguity, are supplemented to ensure that the generated quantitative calculation code accurately expresses the calculation logic. The domain-specific knowledge base also includes a knowledge base on common errors in quantitative calculation code, supplementing it with correct code writing methods for common financial quantitative scenarios to generate accurate and error-prone code. Furthermore, the code agent can integrate the knowledge base of enterprise system interface technical documentation to generate the necessary lines of code for calling relevant internal enterprise interfaces during the code generation process. The code agent dynamically integrates this information and applies engineering methods such as iterative self-correction to generate robust and usable quantitative calculation code for subsequent execution and quantitative testing.

[0071] After generating the quantitative calculation code, the code agent uses the interface agent to obtain professional data and executes the quantitative calculation code to generate quantitative factor data.

[0072] S104. Extract the backtesting method specified in the quantitative research document based on the information extraction agent, call the backtesting tool corresponding to the backtesting method through the interface agent, and use the backtesting tool to perform quantitative backtesting on the quantitative factor data to obtain the backtesting results.

[0073] Quantitative testing is a key step in quantitative research in financial engineering. Quantitative factor testing and quantitative backtesting serve as fundamental methods to validate the correctness and feasibility of the quantitative research ideas and methods presented in the quantitative research document. When conducting quantitative backtesting, the information extraction agent is again used to analyze the quantitative research document and extract the backtesting method for the quantitative backtest. The interface agent then invokes the backtesting tool corresponding to the backtesting method and uses it to perform a quantitative backtest on the quantitative factor data to obtain the backtesting results. For example, if the research validation information in the quantitative research document belongs to a common backtesting strategy, the interface agent invokes the backtesting tool corresponding to the aforementioned backtesting method and performs a quantitative backtest on the quantitative factor data based on real market data to obtain the backtesting results. For example, the MiKuang backtester is invoked to perform a backtest and generate backtesting metrics (such as annualized return and Sharpe ratio). If the research validation information in the quantitative research document belongs to a user-defined backtesting strategy, the code agent generates the corresponding backtesting code. This backtesting code is then used to perform a quantitative backtest on the quantitative factor data to obtain the backtesting results.

[0074] S105. Input the backtest results and the quantitative research results extracted by the information extraction agent from the quantitative research document into the verification agent to obtain the quantitative research conclusion drawn by the verification agent by comparing the backtest results with the quantitative research results.

[0075] Finally, the quantitative research results extracted by the information extraction agent from the quantitative research document are qualitatively and quantitatively compared with the actual backtest results generated through the previous steps to obtain the quantitative research conclusions, that is, to determine whether the quantitative research ideas and conclusions proposed in the quantitative research document are correct.

[0076] Furthermore, the evaluation and optimization agent can provide optimization suggestions for this quantitative research, such as the design direction of quantitative factors and parameter optimization. This provides users with AI-based research optimization and ideas, expands their research ideas, and helps them refine their optimization theory. If the user accepts, the AI-based suggestions can be integrated with the current information and data to create a new quantitative research document, re-execute the entire quantitative research process, and obtain new conclusions and suggestions. This allows for continuous iterative optimization of financial engineering quantitative theory research until user satisfaction is achieved.

[0077] In summary, the present invention provides an intelligent financial engineering quantitative research and reproduction method based on a large model intelligent agent, aiming to provide enterprises and practitioners with an efficient, professional and intelligent financial quantitative virtual research assistant (the virtual research assistant here is an integration of the above-mentioned information extraction intelligent agent, interface intelligent agent, code intelligent agent and verification intelligent agent, which organically connects the above-mentioned intelligent agents through workflows to automatically execute each step of quantitative research). It can interact with users in natural language, receive research tasks issued by users in the form of quantitative research documents, intelligently select appropriate target workflows, and gradually complete the steps required for financial engineering quantitative research according to the target workflow, and finally achieve the task from task description to research conclusions at the data and code levels. Please refer to Figure 2 , Figure 2 This is the second flowchart of the intelligent financial engineering quantitative research and replication method provided by the present invention. The main process of conducting financial engineering quantitative research includes: combining the workflow template and metadata extracted from the quantitative research document by the information extraction agent to clarify the target workflow; when executing the target workflow, the information extraction agent extracts quantitative calculation information from the document, including quantitative research logic; the code agent generates quantitative calculation code based on the quantitative calculation information and data source information, and executes the quantitative calculation code to obtain quantitative factor data; quantitative backtesting of the quantitative factor data using the backtesting tool corresponding to the backtesting method specified in the document to obtain backtesting results; and the verification agent compares the backtesting results with the conclusion information extracted from the document by the information extraction agent to draw quantitative research conclusions. During this process, the interface agent provides support with professional data and calculation tools.

[0078] The intelligent financial engineering quantitative research and replication method provided by the present invention no longer requires professionals to spend a lot of time and energy preparing data and writing code, allowing practitioners to focus on more core work such as quantitative involvement and strategy design, thereby expanding the work efficiency and ability boundaries of human employees.

[0079] In summary, the intelligent financial engineering quantitative research and replication method provided by this invention leverages the powerful semantic understanding, logical reasoning, and code generation capabilities of a large-scale intelligent agent. Once a user provides a quantitative research document, the entire research process, from data to quantitative calculation code, to quantitative backtesting, and conclusion drawing, can be automatically implemented. This improves the efficiency and feasibility of financial engineering quantitative research, lowers the industry barrier, and enables ordinary practitioners to handle complex quantitative research tasks and implement and verify their own quantitative research ideas. It also frees up practitioners to apply more energy to other creative, guiding, and strategic work, enhancing their own value and helping companies optimize talent allocation.

[0080] Based on the above embodiment:

[0081] As an optional embodiment, the information extraction agent extracts quantitative calculation information from the quantitative research document, including:

[0082] The information extraction agent sets prompt words related to the quantitative calculation information in a step-by-step manner from macro to detail, and extracts the quantitative calculation information from the quantitative research document according to each prompt word in order from macro to detail; and / or, the information extraction agent sets prompt words related to the quantitative calculation information from multiple preset dimensions, and extracts the quantitative calculation information from the quantitative research document according to each prompt word.

[0083] The information extraction agent is key to successfully completing the overall quantitative research task. In this embodiment, the information extraction agent employs multiple information extraction techniques. First, prompts related to quantitative calculation information are assigned from a macroscopic to a detailed level, allowing the agent to extract quantitative calculation information from the quantitative research document in a step-by-step, progressive manner. For example, at the macro level, the agent first extracts the summary document theme and core elements from the quantitative research document, and then, at the detailed level, extracts detailed information related to the core elements based on the core elements. Second, prompts related to quantitative calculation information are assigned from multiple preset dimensions, allowing the agent to extract more comprehensive and robust information from the quantitative research document through a multi-angle extraction approach. For example, information is extracted from multiple perspectives related to the core elements of the quantitative research document, verifying and corroborating each other; computational logic descriptions are extracted from the quantitative research document from the perspective of factor construction, and factor computational logic descriptions are extracted from the perspective of computational paragraphs within the document.

[0084] Please refer to Figure 3 , Figure 3 This is a workflow diagram of an information extraction agent provided by the present invention. Figure 3 For example, first use the topic description extraction as a prompt word to extract the research topic and core idea from the document; then use the research object extraction as a prompt word to extract the main research object in the document from the content related to the research topic and core idea; then use the calculation logic detail extraction as a prompt word to extract the calculation logic details from the content related to the main research object in the document.

[0085] Furthermore, as described above, the entire financial engineering quantitative research process requires the use of information extraction agents at multiple points in the process to extract information from quantitative research documents, such as quantitative calculation information, backtesting methods, and quantitative research results. Each node can utilize the step-by-step, multi-angle extraction approach described in this embodiment to extract the required information, simply by adjusting the prompt words accordingly.

[0086] On this basis, the information extraction agent can simultaneously extract quantitative calculation information from quantitative research documents based on each prompt word using two or more different types of large models. The quantitative calculation information extracted from the quantitative research documents by each large model is then deduplicated and integrated to obtain the final quantitative calculation information.

[0087] When using an information extraction agent to extract information from quantitative research documents, multiple large models of different types can be used to extract quantitative computational information from each document. For example, a large model that excels at logical reasoning and a large model that excels at long text can be used to extract quantitative computational information from each document. For example, two or three state-of-the-art large models can be used to extract quantitative computational information from each document. This approach can, to some extent, address the problem of large model hallucinations.

[0088] Furthermore, when the discrepancies between the quantitative calculation information extracted by each large model reach a preset threshold, a prompt is generated to prompt the user to correct the quantitative calculation information. If the discrepancies between the quantitative calculation information extracted by different large models are significant (for example, by determining whether the discrepancies between the quantitative calculation information exceed a preset threshold), user intervention can be introduced. After the user resolves the discrepancies between the quantitative calculation information, the subsequent steps of the quantitative research can be continued. This approach also incorporates user interaction into the quantitative research process, avoiding a completely black-box mechanism and giving users a sense of participation and control.

[0089] Furthermore, after extracting quantitative calculation information from the quantitative research document according to each prompt word, the information extraction agent can also verify the completeness of the extracted quantitative calculation information, that is, to determine whether the currently extracted quantitative calculation information is sufficiently complete and effective and can meet the requirements of subsequent quantitative calculations. Figure 3 , Figure 3 This is a workflow diagram of an information extraction agent provided by the present invention. Specifically, first determine whether the currently extracted quantitative calculation information meets the preset completeness verification conditions. The preset completeness verification conditions are the conditions corresponding to the information required by the code agent to generate quantitative calculation codes when the quantitative calculation information meets the code agent. If so, then enter the step of inputting the quantitative calculation information into the code agent so that the code agent generates quantitative calculation codes based on the quantitative calculation information and the data source information obtained by calling the professional domain knowledge base, data warehouse and professional calculation tools through the interface agent. If not, then enter the step of extracting quantitative calculation information from the quantitative research document according to each prompt word, and continue to extract quantitative calculation information until the quantitative calculation information meets the completeness requirements. For example, the information extraction agent starts from the quantitative logic in accordance with the first principles, and generates a forward and backward dependent, self-consistent and rigorous calculation logic plan, thereby completely extracting the quantitative calculation information for the subsequent code agent to generate professional and rigorous quantitative calculation codes.

[0090] The following describes the process of generating quantitative calculation code by the code agent.

[0091] As an optional embodiment, the code agent generates quantitative calculation code based on the quantitative calculation information and the data source information obtained by the interface agent by calling the professional domain knowledge base, data warehouse, and professional calculation tools, including:

[0092] The code agent generates quantitative calculation code according to the quantitative calculation information and preset prompt words, and calls the professional field knowledge base through the interface agent when generating the quantitative calculation code to adjust the quantitative calculation code in combination with the content in the professional field knowledge base.

[0093] The code agent calls the data warehouse through the interface agent to obtain the relevant data required for sample testing of the quantitative calculation code, and calls professional calculation tools through the interface agent to test the quantitative calculation code using test samples.

[0094] When the quantitative calculation code passes the sample test, the quantitative calculation code is stored in the database as the final quantitative calculation code; when the quantitative calculation code fails the sample test, the error and abnormal information that occurs during the execution of the quantitative calculation code are compiled into the professional domain knowledge base, and the quantitative calculation code is generated again according to the preset prompt words based on the quantitative calculation information, and the professional domain knowledge base is called through the interface intelligent body when generating the quantitative calculation code to adjust the steps of the quantitative calculation code in combination with the content in the professional domain knowledge base.

[0095] Please refer to Figure 4 , Figure 4 This is a workflow diagram of a code agent provided by the present invention. In order to generate professional and format-compliant quantitative calculation codes, preset prompt words are pre-set, including but not limited to code generation requirements, output code format, and domain knowledge booth information. The code agent generates quantitative calculation codes according to the preset prompt words based on the quantitative calculation information, that is, Figure 4 The LLM (Large Language Model Code) code in .

[0096] In order to ensure the correctness and executability of the quantitative calculation code, the present invention also sets up a professional field knowledge base. Specifically, the present invention pre-collects general information in the field of financial quantification based on RAG (Retrieval-Augmented Generation) and stores it in the professional field knowledge base. When the code agent generates the quantitative calculation code, it will use the interface agent to preferentially query the relevant professional information from the professional field knowledge base, thereby generating a quantitative calculation code that meets the industry requirements. At the same time, several cases in which code errors are prone to occur in financial data processing and quantitative calculation scenarios are also collected and stored in the professional field knowledge base. These cases are used as supplementary information to assist the code agent in generating correct and effectively executable quantitative calculation codes, thereby avoiding common errors when the code agent generates quantitative calculation codes. The professional field knowledge base also includes technical documentation of enterprise private domain data and calculation tool interfaces, which are used to integrate calls to enterprise private domain data or business interfaces into the generated quantitative calculation code, such as obtaining relevant data from the enterprise data warehouse.

[0097] In addition, the professional domain knowledge base also stores information such as error correction suggestions for abnormal errors that occur during the execution of quantitative computing codes, so that a self-feedback iteration mechanism can be used to allow the code agent to self-diagnose and repair the current problem code, ensuring the executability and correctness of the quantitative computing code.

[0098] The present invention further utilizes the interface intelligent body to call the data in the data warehouse and professional computing tools, and uses the test sample to test the generated quantitative computing code, test the actual execution status of the quantitative computing code, and verify the correctness of the execution result of the quantitative computing code from aspects such as format and type. When the quantitative computing code passes the sample test, it is considered that the quantitative computing code has passed the test and the quantitative computing code is saved; when the quantitative computing code fails the sample test, the self-feedback iteration mechanism is used to adjust the quantitative computing code again according to the problems that arise, and the sample test is re-performed on the adjusted quantitative computing code until the quantitative computing code passes the sample test. The quantitative computing code that passes the sample test is saved as the final quantitative computing code, further ensuring the correctness of the quantitative computing code.

[0099] As an optional embodiment, after the code agent executes the quantization calculation code to generate quantization factor data, the method further includes:

[0100] The information extraction agent extracts the backtesting method specified in the quantitative research document, and calls the backtesting tool corresponding to the backtesting method through the interface agent. The backtesting tool is used to perform quantitative testing on the quantitative factor data in combination with the current real market data to obtain the current test results, so as to determine the performance of the quantitative factor data in the current market.

[0101] This embodiment takes into account that quantitative testing is based on historical data, and the quantitative ideas that were effective in the past may be invalid at present. It is very important to test whether the quantitative research is applicable to the current market environment. Therefore, the present invention supports users to deploy the quantitative research results they need to the managed simulation with one click. Please refer to Figure 2 , Figure 2 This is the second flow chart of the intelligent financial engineering quantitative research and replication method provided by the present invention. This method can intelligently screen effective quantitative factor data for real-time hosting and simulation. For example, the current performance of quantitative factors can be evaluated based on real-time data. This involves utilizing the backtesting tools corresponding to the backtesting methods specified in the quantitative research document, combined with current real market data, to conduct new quantitative tests on the quantitative factor data. This allows users to obtain the performance of the quantitative research under current market conditions, more effectively assisting users in applying quantitative research in real trading.

[0102] As an optional embodiment, the interface agent calls a professional domain knowledge base, a data warehouse, professional computing tools, and a backtesting tool, including:

[0103] Each first alternative tool in the first layer of the target access object is used as the current alternative tool, and each current alternative tool is written into the function call parameter of the interface agent so that the interface agent can select the current target tool from the current alternative tools; wherein, the target access object includes a professional domain knowledge base, a data warehouse, a professional calculation tool and a backtesting tool, and the professional domain knowledge base, the data warehouse, the professional calculation tool and the backtesting tool are all reconstructed into a tree structure in advance according to a preset hierarchical standard.

[0104] The alternative tools in the next layer of the node where the current target tool is located are used as new current alternative tools, and the step of writing each current alternative tool into the function call parameter of the interface agent is entered so that the interface agent can select the current target tool from the current alternative tools. Until the current target tool is the actual tool, the structured parameters required to call the actual tool are returned to the interface agent so that the interface agent can call the target access object to complete the corresponding function.

[0105] In the present invention, generating quantitative calculation code, executing quantitative calculation code to generate quantitative factor data, and performing quantitative backtesting all require the support of an interface intelligent agent, and it is necessary to call a professional domain knowledge base, a data warehouse, professional calculation tools, and a backtesting tool through the interface intelligent agent.

[0106] In this embodiment, the professional domain knowledge base, data warehouse, professional calculation tools and backtesting tools are used as target access objects, and the professional domain knowledge base, data warehouse, professional calculation tools and backtesting tools are all reconstructed into a tree structure in advance according to the preset hierarchical standards. This requires human experts to classify the existing tools in layers in advance, which is equivalent to giving the intelligent agent expert opinions. It is suitable for use in tool intelligent call scenarios with complex business scenarios and long business chains. The target access objects can be hierarchically classified according to partial systems, functional granularity, etc., and each node in the abstract class layer needs to build a category tool (the function of the tool is to return the position of the current node) until the actual tool of the last layer. The tool consists of two parts, one is the text description of the tool, including the actual function of the tool, the tool name, the parameters contained in the tool, and the parameter type and parameter meaning; the second is the implementation of the tool, including the acquisition of tool parameters, which can be processed into the required parameters through the general parameter extractor. The tool with the build number is registered in the project for subsequent call (i.e., written to the variable space).

[0107] The interface agent provided in this embodiment essentially accesses the target object by making multiple queries, each query returning only the key information at that layer, gradually accessing the actual tools within the target object. This design overcomes the limitations of large language models, such as their inability to handle a large number of tools at once, and the context length limitations.

[0108] Please refer to Figure 5 , Figure 5 The present invention provides a workflow diagram of an interface agent. Specifically, the interface agent accesses downward from the top level of the tree structure, firstly accessing the first candidate tools (i.e. Figure 5 The tool corresponding to the first-level classification in the interface is used as the current candidate tool, and each current candidate tool is written into the function call parameter of the interface agent so that the interface agent can select the current target tool from the current candidate tools. Then, each candidate tool in the next layer of the node where the current target tool is located is used as the new current candidate tool, and the step of writing each current candidate tool into the function call parameter of the interface agent is entered so that the interface agent can select the current target tool from the current candidate tools. When the current target tool is the actual tool, the structured parameters required for calling the actual tool are returned to the interface agent so that the interface agent can call the target access object to complete the corresponding function.

[0109] Generally speaking, the interface agent first forks to the next level based on the overall intent. The next level then forks to the second level based on the detailed intent. This classification strategy continues until the final actual tool is reached, thus constructing a tree-like organizational structure. The interface agent can accurately hit the required actual tool through multiple rounds of requests, smoothly call the target access object, and complete the corresponding function. For example, to obtain the market information of a specific stock for the past ten years, multiple requests are made to hit the stock market query interface of the data warehouse.

[0110] Please refer to Figure 6 , Figure 6 This is a schematic diagram of the structure of an intelligent financial engineering quantitative research and reproduction system provided by the present invention, which includes:

[0111] The target workflow determination unit 601 is configured to utilize an information extraction agent to extract meta-information related to the workflow structure from the quantitative research document provided by the user, and determine and execute the target workflow by combining the meta-information and the workflow template.

[0112] The quantitative calculation information extraction unit 602 is used to input the quantitative research document into the information extraction agent when executing the target workflow, and obtain the quantitative calculation information extracted by the information extraction agent from the quantitative research document.

[0113] The quantization factor data generation unit 603 is used to input the quantization calculation information into the code agent so that the code agent generates the quantization calculation code based on the quantization calculation information and the data source information obtained by calling the professional domain knowledge base, data warehouse and professional calculation tools through the interface agent, obtains professional data through the interface agent and executes the quantization calculation code to generate the quantization factor data.

[0114] The backtesting unit 604 is used to extract the backtesting method specified in the quantitative research document based on the information extraction agent, and call the backtesting tool corresponding to the backtesting method through the interface agent, and use the backtesting tool to perform quantitative backtesting on the quantitative factor data to obtain the backtesting results.

[0115] The verification unit 605 is used to input the backtest results and the quantitative research results extracted from the quantitative research document by the information extraction agent into the verification agent, and obtain the quantitative research conclusion drawn by the verification agent by comparing the backtest results with the quantitative research results.

[0116] For a detailed introduction to the intelligent financial engineering quantitative research and reproduction system provided by the present invention, please refer to the above-mentioned embodiment of the intelligent financial engineering quantitative research and reproduction method, and the present invention will not elaborate on it here.

[0117] Please refer to Figure 7 , Figure 7This is a schematic diagram of the structure of an intelligent financial engineering quantitative research and reproduction device provided by the present invention, which includes:

[0118] Memory 701, used for storing computer programs;

[0119] Processor 702 is configured to implement any of the steps of the aforementioned intelligent financial engineering quantitative research and replication method when executing a computer program.

[0120] For a detailed introduction to the intelligent financial engineering quantitative research and reproduction device provided by the present invention, please refer to the above-mentioned embodiment of the intelligent financial engineering quantitative research and reproduction method, and the present invention will not elaborate on it here.

[0121] In order to solve the above technical problems, the present invention also provides a storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of any of the above-mentioned intelligent financial engineering quantitative research and reproduction methods are implemented.

[0122] For a detailed introduction to the storage medium provided by the present invention, please refer to the above-mentioned embodiments of the intelligent financial engineering quantitative research and reproduction method, and the present invention will not elaborate on them here.

[0123] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. It should also be noted that in this specification, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. Moreover, the terms "comprises", "includes" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, article or device that includes a series of elements includes not only those elements, but also other elements that are not explicitly listed, or also includes elements that are inherent to such process, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a..." does not exclude the presence of other identical elements in the process, article or device that includes the element.

[0124] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An intelligent financial engineering quantitative research and reproduction method, characterized by: include: Utilize an information extraction agent to extract meta-information related to the workflow structure from the quantitative research document provided by the user, and determine and execute the target workflow by combining the meta-information with the workflow template; When executing the target workflow, a quantitative research document is input into an information extraction agent, and quantitative calculation information extracted by the information extraction agent from the quantitative research document is obtained, wherein the quantitative calculation information includes a description of the quantitative factor calculation logic, a calculation formula, a step-by-step calculation autonomous plan, data source information, and factor structured meta-information; Inputting the quantitative calculation information into a code agent so that the code agent generates quantitative calculation code based on the quantitative calculation information and data source information obtained by calling a professional domain knowledge base, a data warehouse, and professional calculation tools through an interface agent, obtains professional data through the interface agent, and executes the quantitative calculation code to generate quantitative factor data; Extracting the backtesting method specified in the quantitative research document according to the information extraction agent, calling the backtesting tool corresponding to the backtesting method through the interface agent, and performing quantitative backtesting on the quantitative factor data using the backtesting tool to obtain a backtesting result; Inputting the backtest results and the quantitative research results extracted from the quantitative research document by the information extraction agent into the verification agent, and obtaining the quantitative research conclusion drawn by the verification agent by comparing the backtest results with the quantitative research results; The code agent generates a quantitative calculation code based on the quantitative calculation information and the data source information obtained by the interface agent calling a professional domain knowledge base, a data warehouse, and a professional calculation tool, including: The code agent generates a quantitative calculation code according to the quantitative calculation information and a preset prompt word, and calls a professional domain knowledge base through the interface agent when generating the quantitative calculation code to adjust the quantitative calculation code in combination with the content of the professional domain knowledge base; The code agent calls the data warehouse through the interface agent to obtain relevant data required for sample testing of the quantitative calculation code, and calls the professional calculation tool through the interface agent to test the quantitative calculation code using the test sample; When the quantitative calculation code passes the sample test, the quantitative calculation code is stored in a library as the final quantitative calculation code; When the quantitative calculation code fails the sample test, the error and exception information that occurs during the execution of the quantitative calculation code will be compiled into the professional field knowledge base, and the process of generating the quantitative calculation code according to the preset prompt words based on the quantitative calculation information will be re-entered, and when generating the quantitative calculation code, the professional field knowledge base will be called through the interface intelligent body to adjust the quantitative calculation code in combination with the content in the professional field knowledge base.

2. The intelligent financial engineering quantitative research and reproduction method according to claim 1, characterized in that: The information extraction agent extracts quantitative calculation information from the quantitative research document, including: The information extraction agent sets prompt words related to the quantitative calculation information in a step-by-step manner from macro to detail, and extracts the quantitative calculation information from the quantitative research document according to the prompt words in order from macro to detail; And / or, the information extraction agent sets prompt words related to the quantitative calculation information from a plurality of preset dimensions, and extracts the quantitative calculation information from the quantitative research document according to each of the prompt words.

3. The intelligent financial engineering quantitative research and reproduction method according to claim 2, characterized in that: The information extraction agent extracts quantitative calculation information from the quantitative research document according to each of the prompt words, including: The information extraction agent uses two or more different types of large models to extract quantitative calculation information from the quantitative research document respectively; The quantitative calculation information extracted by each of the large models from the quantitative research document is deduplicated and integrated to obtain the final quantitative calculation information, and when the difference between the quantitative calculation information extracted by each large model reaches a preset threshold, a prompt message is generated to prompt the user to correct the quantitative calculation information.

4. The intelligent financial engineering quantitative research and reproduction method according to claim 2, characterized in that: After the information extraction agent extracts the quantitative calculation information from the quantitative research document according to each of the prompt words, the information extraction agent further includes: Determining whether the currently extracted quantitative calculation information satisfies a preset completeness verification condition, where the preset completeness verification condition is a condition corresponding to when the quantitative calculation information satisfies information required by the code agent to generate the quantitative calculation code; If so, the step of inputting the quantitative calculation information into the code agent is performed so that the code agent generates quantitative calculation code based on the quantitative calculation information and the data source information obtained by calling the professional domain knowledge base, data warehouse and professional calculation tools through the interface agent; If not, the process proceeds to the step of extracting the quantitative calculation information from the quantitative research document according to each of the prompt words.

5. The intelligent financial engineering quantitative research and reproduction method according to claim 1, characterized in that: After the code agent executes the quantization calculation code to generate quantization factor data, the method further includes: The information extraction agent extracts the backtesting method specified in the quantitative research document, and calls the backtesting tool corresponding to the backtesting method through the interface agent. The backtesting tool is used to perform quantitative testing on the quantitative factor data in combination with the current real market data to obtain the current test results, so as to determine the performance of the quantitative factor data in the current market.

6. The intelligent financial engineering quantitative research and reproduction method according to claim 1, characterized in that: Through the interface agent, professional domain knowledge base, data warehouse, professional calculation tools and backtesting tools are called, including: Taking each first candidate tool in the first layer of the target access object as a current candidate tool, and writing each current candidate tool into a function call parameter of the interface agent so that the interface agent selects the current target tool from each current candidate tool; The target access objects include the professional domain knowledge base, the data warehouse, the professional computing tools, and the backtesting tools, and the professional domain knowledge base, the data warehouse, the professional computing tools, and the backtesting tools are all reconstructed into a tree structure in advance according to a preset hierarchical standard; The alternative tools in the next layer of the node where the current target tool is located are used as new current alternative tools, and the step of writing each current alternative tool into the function call parameter of the interface agent is entered so that the interface agent can select the current target tool from the current alternative tools. Until the current target tool is the actual tool, the structured parameters required to call the actual tool are returned to the interface agent so that the interface agent can call the target access object to complete the corresponding function.

7. An intelligent financial engineering quantitative research and reproduction system, characterized by: include: a target workflow determination unit, configured to extract meta-information related to the workflow structure from the quantitative research document provided by the user using an information extraction agent, and determine and execute the target workflow by combining the meta-information with the workflow template; a quantitative calculation information extraction unit, configured to input a quantitative research document into an information extraction agent when executing the target workflow, and obtain quantitative calculation information extracted by the information extraction agent from the quantitative research document, wherein the quantitative calculation information includes a description of the quantitative factor calculation logic, a calculation formula, a step-by-step calculation autonomous plan, data source information, and factor structured meta-information; a quantization factor data generation unit, configured to input the quantization calculation information into a code agent, so that the code agent generates quantization calculation code based on the quantization calculation information and data source information obtained by calling a professional domain knowledge base, a data warehouse, and professional calculation tools through an interface agent, obtain professional data through the interface agent, and execute the quantization calculation code to generate quantization factor data; A backtesting unit, configured to extract the backtesting method specified in the quantitative research document based on the information extraction agent, call the backtesting tool corresponding to the backtesting method through the interface agent, and use the backtesting tool to perform quantitative backtesting on the quantitative factor data to obtain a backtesting result; a verification unit, configured to input the backtest results and the quantitative research results extracted from the quantitative research document by the information extraction agent into a verification agent, and obtain a quantitative research conclusion drawn by the verification agent by comparing the backtest results with the quantitative research results; The code agent generates a quantitative calculation code based on the quantitative calculation information and the data source information obtained by the interface agent calling a professional domain knowledge base, a data warehouse, and a professional calculation tool, including: The code agent generates a quantitative calculation code according to the quantitative calculation information and a preset prompt word, and calls a professional domain knowledge base through the interface agent when generating the quantitative calculation code to adjust the quantitative calculation code in combination with the content of the professional domain knowledge base; The code agent calls the data warehouse through the interface agent to obtain relevant data required for sample testing of the quantitative calculation code, and calls the professional calculation tool through the interface agent to test the quantitative calculation code using the test sample; When the quantitative calculation code passes the sample test, the quantitative calculation code is stored in a library as the final quantitative calculation code; When the quantitative calculation code fails the sample test, the error and exception information that occurs during the execution of the quantitative calculation code will be compiled into the professional field knowledge base, and the process of generating the quantitative calculation code according to the preset prompt words based on the quantitative calculation information will be re-entered, and when generating the quantitative calculation code, the professional field knowledge base will be called through the interface intelligent body to adjust the quantitative calculation code in combination with the content in the professional field knowledge base.

8. An intelligent financial engineering quantitative research and reproduction device, characterized by: include: memory for storing computer programs; A processor, configured to implement the steps of the intelligent financial engineering quantitative research and reproduction method as described in any one of claims 1 to 6 when executing the computer program.

9. A storage medium, characterized in that: The storage medium stores a computer program, which, when executed by a processor, implements the steps of the intelligent financial engineering quantitative research and reproduction method according to any one of claims 1 to 6.

Images