AI autonomous quantitative investment application technology system

Abstract

The invention relates to the technical field of AI investment application, in particular to an AI autonomous quantitative investment application technology system, which comprises information data collection, data representation, AI autonomous learning investment strategy, simulation investment of AI in an actual market environment, profit condition favorite system output and a reestablishment system under a loss condition. According to the application, data information associated with market investment can be fully collected, related policies can be effectively collected and interpreted, price trends of investment products can be fully collected, and corresponding data, information and policies are associated to understand the influence of market information and policies on the price trends of the products, so that the product price trends can be quickly and accurately identified. And lagging of policies or information and corresponding emergencies can be considered, establishment of the autonomous investment technology is completed through an AI intelligent learning technique, simulation operation is performed in a real environment, and whether the autonomous investment technology is qualified or not is judged.

Description

Technical Field

[0001] This invention relates to the field of AI investment application technology, and in particular to an AI-based autonomous quantitative investment application technology system. Background Technology

[0002] With the development of artificial intelligence technology, the financial investment field is increasingly reliant on intelligent decision support systems. Traditional investment strategies often depend on static data analysis and experience-based judgment, making it difficult to adapt to rapid market changes and the complexity of investor behavior. Prospect theory and cognitive biases, as important theories in behavioral economics, offer new perspectives for understanding and predicting investor decision-making behavior. However, applying these theories to the optimization of actual investment strategies requires an intelligent system capable of integrating multi-source data, dynamically learning, and adjusting strategies.

[0003] There is an urgent need for a system that can integrate prospect theory and cognitive biases and optimize investment strategies in real time through machine learning technology.

[0004] A method and system for assisting investment strategies, disclosed in CN119919238A, first collects and integrates market data to obtain a training set; then, it develops a data analysis model using a neural network algorithm, and trains the model using the training set and manually labeled investment strategy assistance results; finally, it obtains input data from the target user based on a pre-established user interface, processes the input data using the trained data analysis model, and obtains the target user's investment strategy assistance results. This application can optimize investment strategies in real time through machine learning technology.

[0005] The above-mentioned technical solutions cannot adequately summarize and process market, policy, and various information, nor can they comprehensively and independently analyze their impact on market prices, and therefore cannot effectively carry out independent investment operations. Therefore, improvements are needed. Summary of the Invention

[0006] The purpose of this invention is to address the shortcomings of existing technologies by proposing an AI-based autonomous quantitative investment application technology system.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] The AI-driven autonomous quantitative investment application technology system includes the following steps:

[0009] S1. Collect market data, information, and policies;

[0010] S2. Output market price fluctuations separately and create a price fluctuation line chart;

[0011] S3. Output other market data information separately in sequence and create line charts. At the same time, display the price line chart and other line charts on the same line chart.

[0012] S4. Mark the policies and international information according to the timeline;

[0013] S5. It can autonomously learn and determine the impact of market data, policies, and international information on prices; it can also collect investment operations from multiple investors to enable AI to learn autonomously.

[0014] S6. Simulated investment training: Set profit targets, conduct independent investment training, and establish a complete system; and set up a corresponding risk system to automatically control risks;

[0015] S7. When profitable, output the system solution;

[0016] S8. When there is a loss, return to S1 to rebuild the system plan.

[0017] Compared with existing technologies, this application can fully collect market investment-related data and information, effectively collect and interpret relevant policies, and fully collect the price trends of investment products. It also links the relevant data, information and policies to understand the impact of market information and policies on product price trends. Furthermore, it can consider the lag of policies or information and corresponding emergencies, and use AI to intelligently learn and establish autonomous investment technology. At the same time, it conducts simulated operations in a real environment to determine its qualification.

[0018] Preferably, the market data includes historical price data, the number of product manufacturers, product inventory, wage changes, historical volatility, GDP growth rate, unemployment rate, and market sentiment index; the information and policies include the collection of policy information and the collection of international information.

[0019] Furthermore, in practice, taking the stock market as an example, to fully collect information within the stock market, we can first conduct AI autonomous learning based on a major category, such as the medical industry, the sports product industry, or the apparel industry. We can fully collect all relevant information for that industry, as well as related stock information and the information data behind the companies listed on those stocks. This includes information on company personnel, personnel changes, average and median wages, social security information, factory location information, and other information related to company operations and senior management. We can also fully obtain information on relevant policies for the industry in which the company operates and the situation of upstream and downstream related industries, as well as relevant international industry information and policies.

[0020] Preferably, the step of marking policy and international information according to the timeline includes summarizing market policy information, determining its correlation with price fluctuations, and marking it on the corresponding timeline, and summarizing international information, determining its correlation with price fluctuations, and marking it on the corresponding timeline.

[0021] Furthermore, the acquired information will be organized, primarily using price fluctuation line charts, while also considering the lag in information and policies, as well as the suddenness of related black swan events.

[0022] Preferably, the collection of market data, information, and policies in step S1 includes the following steps:

[0023] Data source identification: Identify the sources of data to be collected, including financial databases, market data providers, news reports, and social media platforms;

[0024] Data scraping: Using web crawling technology to scrape relevant financial data from a defined data source;

[0025] Data cleaning: Using unsupervised learning algorithms to clean the collected raw data that contains noise, duplication, or irrelevant information, removing outliers, filling in missing values, and standardizing the data.

[0026] Data integration: Integrating cleaned data according to a specific format or structure to facilitate subsequent analysis and processing;

[0027] Data storage: Store the integrated data in an appropriate database or data warehouse for subsequent data analysis and optimization.

[0028] Furthermore, it is essential to ensure that the collected data is comprehensively organized and optimized, and to fully extract effective information. Because the data is complex, it is necessary to extract relevant information from the complex data and effectively determine the priority of information association.

[0029] Preferably, the autonomous learning and judgment in S5 adopts a neural network algorithm to develop a data analysis model, and the data analysis model is trained by the training set and manually labeled investment strategy auxiliary results; wherein, the data analysis model is used to analyze the decision-making patterns of investors under different market conditions, and the data analysis model includes at least probability weighting and value functions;

[0030] Specifically, the auxiliary result NewOutput for the investment strategy is determined by NewOutput = Model(x, p, θuser), where Model is a data analysis model; x and p represent the characteristics and relevant probabilities of the investment plan, respectively; and θuser represents the set of parameters adjusted by the user, which can incorporate the influence of risk preference and time.

[0031] Furthermore, during training, the collected operational information from operators is input into the training content to understand how they operate and make judgments when they are profitable; how they reduce losses when they are losing; and how they consider giving up when they are holding steady.

[0032] The above solution can effectively enable AI to learn fully and integrate the operating techniques of operators, thus enabling better autonomous investment.

[0033] Preferably, the risk system in S6 includes the following steps:

[0034] A module for generating real-time risk monitoring: This module collects real-time market data, assesses whether market price fluctuations meet expected prices, and determines whether profit and loss situations meet investment expectations.

[0035] Risk assessment and analysis: Utilizing machine learning algorithms and techniques to process and analyze collected data, assess the current risk level of the investment portfolio, and identify potential risk factors;

[0036] Set risk thresholds: Based on the user's risk preferences and investment goals, set reasonable risk thresholds as the standard for triggering risk control mechanisms;

[0037] Develop risk control strategies: Once the risk exceeds the set threshold, the risk control module will be automatically triggered, generating corresponding risk control measures, such as increasing holdings, reducing holdings, or selling off, based on pre-set strategies and algorithms.

[0038] Implement risk control measures: The system automatically adjusts the portfolio allocation based on the generated risk control measures to reduce the risk level.

[0039] Furthermore, AI learning algorithms are used to analyze the collected data, assess the current risk level, identify potential risk factors, and adjust investment allocation according to the risk situation.

[0040] Preferably, the risk control module performs risk assessment using the logistic regression algorithm in machine learning. This algorithm, by mining and analyzing large amounts of data, helps the system more accurately identify potential risks and formulate corresponding risk control strategies. The formula is as follows: P(y=1|x) represents the probability that the output variable y is 1 given the input variable x. w is the weight vector, x is the input variable vector, b is the bias term, and e is the base of the natural logarithm. When setting a risk threshold, y=1 is interpreted as the occurrence of a risk event, i.e., the portfolio loss exceeds a certain threshold, while y=0 is interpreted as the risk event not occurring. By training the logistic regression algorithm, we can learn the optimal values ​​of the weight vector w and the bias term b, enabling the investment optimization model to accurately predict the probability of risk events. This helps us understand the model's predictive ability and adjust the model parameters accordingly to automatically adjust the portfolio allocation to reduce the risk level.

[0041] Furthermore, the system monitors the risk status of the investment portfolio in real time and adjusts investment strategies promptly according to market changes to keep the risk within a controllable range. This module can set a risk threshold, and once the threshold is exceeded, the system will automatically trigger the risk control mechanism and take corresponding risk control measures.

[0042] The beneficial effects of this invention are:

[0043] 1. Organize the acquired information, primarily using price fluctuation line charts, while also considering the lag in information and policy, as well as the suddenness of related black swan events; ensure that the collected data is comprehensively organized and optimized, fully extracting effective information. Due to the complexity of the data, it is necessary to extract relevant information from the complex data and effectively determine the priority of information correlation.

[0044] 2. During training, input the collected operational information from operators into the training content to understand how they operate and judge when they are profitable; how they reduce losses when they are losing; and how they consider giving up when they are holding steady.

[0045] The above solution can effectively enable AI to learn fully and integrate the operating techniques of operators, thus enabling better autonomous investment.

[0046] 3. Utilize AI learning algorithms to analyze collected data, assess the current risk level, identify potential risk factors, and adjust investment allocation according to the risk situation;

[0047] 4. Monitor the risk status of the investment portfolio in real time and adjust investment strategies in a timely manner according to market changes to keep the risk within a controllable range. This module can set a risk threshold. Once the threshold is exceeded, the system will automatically trigger the risk control mechanism and take corresponding risk control measures. Attached Figure Description

[0048] Figure 1 This is an application flowchart of the AI ​​autonomous quantitative investment application technology system proposed in this invention;

[0049] Figure 2 This is an information and policy framework diagram of the AI ​​autonomous quantitative investment application technology system proposed in this invention;

[0050] Figure 3 This is a flowchart illustrating the AI-driven autonomous quantitative investment application technology system proposed in this invention, which annotates policies and international information according to a timeline. Detailed Implementation

[0051] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0052] Reference Figure 1-3 The AI-driven autonomous quantitative investment application technology system includes the following steps:

[0053] S1. Collect market data, information, and policies;

[0054] S2. Output market price fluctuations separately and create a price fluctuation line chart;

[0055] S3. Output other market data information separately in sequence and create line charts. At the same time, display the price line chart and other line charts on the same line chart.

[0056] S4. Mark the policies and international information according to the timeline;

[0057] S5. It can autonomously learn and determine the impact of market data, policies, and international information on prices; it can also collect investment operations from multiple investors to enable AI to learn autonomously.

[0058] S6. Simulated investment training: Set profit targets, conduct independent investment training, and establish a complete system; and set up a corresponding risk system to automatically control risks;

[0059] S7. When profitable, output the system solution;

[0060] S8. When there is a loss, return to S1 to rebuild the system plan.

[0061] In this invention, market data includes historical price data, the number of product manufacturers, product inventory, wage changes, historical volatility, GDP growth rate, unemployment rate, and market sentiment index; information and policies include the collection of policy information and international information; in practice, taking the stock market as an example, to fully collect information within the stock market, AI can first conduct autonomous learning based on a major category, such as the medical industry, the sports product industry, or the apparel industry, to fully collect all relevant information for that industry, as well as relevant stock information, and information data behind the companies behind the corresponding stocks, such as company personnel, personnel changes, average and median wages, social security information, factory location information, and other information related to company operations and senior management, and to fully obtain relevant policies for the industry in which the company operates and the situation of upstream and downstream related industries; at the same time, relevant international industry information and policies are also obtained.

[0062] In this invention, policy and international information are annotated according to a timeline, including summarizing market policy information and determining its correlation with price fluctuations, and then annotating it on the corresponding timeline; international information is also summarized and its correlation with price fluctuations is determined, and then annotated on the corresponding timeline; the acquired information is organized, mainly using price fluctuation line charts, while also considering the lag of information and policies, as well as the suddenness of related black swan events.

[0063] In this invention, the collection of market data, information, and policies in S1 includes the following steps:

[0064] Data source identification: Identify the sources of data to be collected, including financial databases, market data providers, news reports, and social media platforms;

[0065] Data scraping: Using web crawling technology to scrape relevant financial data from a defined data source;

[0066] Data cleaning: Using unsupervised learning algorithms to clean the collected raw data that contains noise, duplication, or irrelevant information, removing outliers, filling in missing values, and standardizing the data.

[0067] Data integration: Integrating cleaned data according to a specific format or structure to facilitate subsequent analysis and processing;

[0068] Data storage: Store the integrated data in an appropriate database or data warehouse for subsequent data analysis and optimization;

[0069] Ensure that the collected data is comprehensively organized and optimized, and fully extract effective information. Because the data is complex, it is necessary to extract relevant information from the complex data and effectively determine the priority of information association.

[0070] In this invention, the autonomous learning and judgment in S5 adopts a neural network algorithm to develop a data analysis model, and the data analysis model is trained by training set and manually labeled investment strategy auxiliary results; wherein, the data analysis model is used to analyze the decision-making patterns of investors under different market conditions, and the data analysis model includes at least probability weighting and value function;

[0071] Specifically, the auxiliary result NewOutput for the investment strategy is determined by NewOutput = Model(x, p, θuser), where Model is the data analysis model; x and p represent the characteristics and relevant probabilities of the investment plan, respectively; and θuser represents the set of parameters adjusted by the user, which can incorporate the influence of risk preference and time.

[0072] During training, the collected operational information from operators is input into the training content to understand how they operate and judge when they are profitable; how they reduce losses when they are losing; and how they consider giving up when they are holding steady.

[0073] The above solution can effectively enable AI to learn fully and integrate the operating techniques of operators, thus enabling better autonomous investment.

[0074] In this invention, the risk system in S6 includes the following steps:

[0075] A module for generating real-time risk monitoring: This module collects real-time market data, assesses whether market price fluctuations meet expected prices, and determines whether profit and loss situations meet investment expectations.

[0076] Risk assessment and analysis: Utilizing machine learning algorithms and techniques to process and analyze collected data, assess the current risk level of the investment portfolio, and identify potential risk factors;

[0077] Set risk thresholds: Based on the user's risk preferences and investment goals, set reasonable risk thresholds as the standard for triggering risk control mechanisms;

[0078] Develop risk control strategies: Once the risk exceeds the set threshold, the risk control module will be automatically triggered, generating corresponding risk control measures, such as increasing holdings, reducing holdings, or selling off, based on pre-set strategies and algorithms.

[0079] Implement risk control measures: The system automatically adjusts the portfolio allocation based on the generated risk control measures to reduce the risk level;

[0080] By using AI learning algorithms to analyze the collected data, assess the current risk level, identify potential risk factors, and adjust investment allocation according to the risk situation.

[0081] In this invention, the risk control module uses the logistic regression algorithm in machine learning to assess risk. This algorithm, by mining and analyzing large amounts of data, helps the system more accurately identify potential risks and formulate corresponding risk control strategies. The formula is as follows: P(y=1|x) represents the probability that the output variable y is 1 given the input variable x. w is the weight vector, x is the input variable vector, b is the bias term, and e is the base of the natural logarithm. When setting the risk threshold, y=1 is interpreted as the occurrence of a risk event, i.e., the portfolio loss exceeds a certain threshold, while y=0 is interpreted as the risk event not occurring. By training the logistic regression algorithm, we can learn the optimal values ​​of the weight vector w and the bias term b, enabling the investment optimization model to accurately predict the probability of risk events. This helps us understand the predictive ability of the model and adjust the model parameters accordingly to automatically adjust the portfolio allocation to reduce the risk level.

[0082] The system monitors the risk status of the investment portfolio in real time and adjusts investment strategies promptly according to market changes to keep the risk within a controllable range. This module allows you to set a risk threshold, and once the threshold is exceeded, the system will automatically trigger the risk control mechanism and take corresponding risk control measures.

[0083] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. An AI-driven autonomous quantitative investment application technology system, characterized in that: Includes the following steps: S1. Collect market data, information, and policies; S2. Output market price fluctuations separately and create a price fluctuation line chart; S3. Output other market data information separately in sequence and create line charts. At the same time, display the price line chart and other line charts on the same line chart. S4. Mark the policies and international information according to the timeline; S5. Autonomous learning and judgment of the impact of market data, policies, and international information on prices; Collect investment data from multiple investors to enable AI to learn autonomously. S6. Simulated investment training: Set profit targets, conduct independent investment training, and establish a complete system; and set up a corresponding risk system to automatically control risks; S7. When profitable, output the system solution; S8. When there is a loss, return to S1 to rebuild the system plan.

2. The AI-driven autonomous quantitative investment application technology system according to claim 1, characterized in that: The market data includes historical price data, number of product manufacturers, product inventory, wage changes, historical volatility, GDP growth rate, unemployment rate, and market sentiment index; the information and policies include the collection of policy information and international information.

3. The AI-driven autonomous quantitative investment application technology system according to claim 1, characterized in that: The step of annotating policies and international information according to the timeline includes summarizing market policy information, determining its correlation with price fluctuations, and then annotating it on the corresponding timeline; and summarizing international information, determining its correlation with price fluctuations, and then annotating it on the corresponding timeline.

4. The AI-driven autonomous quantitative investment application technology system according to claim 1, characterized in that: The collection of market data, information, and policies in S1 includes the following steps: Data source identification: Identify the sources of data to be collected, including financial databases, market data providers, news reports, and social media platforms; Data scraping: Using web crawling technology to scrape relevant financial data from a defined data source; Data cleaning: Using unsupervised learning algorithms to clean the collected raw data that contains noise, duplication, or irrelevant information, removing outliers, filling in missing values, and standardizing the data. Data integration: Integrating cleaned data according to a specific format or structure to facilitate subsequent analysis and processing; Data storage: Store the integrated data in an appropriate database or data warehouse for subsequent data analysis and optimization.

5. The AI-driven autonomous quantitative investment application technology system according to claim 1, characterized in that: Therefore, the autonomous learning and judgment in S5 adopts a neural network algorithm to develop a data analysis model, and trains the data analysis model using the training set and manually labeled investment strategy auxiliary results; wherein, the data analysis model is used to analyze the decision-making patterns of investors under different market conditions, and the data analysis model includes at least probability weighting and a value function; Specifically, the auxiliary result NewOutput for the investment strategy is determined by NewOutput = Model(x, p, θuser), where Model is a data analysis model; x and p represent the characteristics and relevant probabilities of the investment plan, respectively; and θuser represents the set of parameters adjusted by the user, which can incorporate the influence of risk preference and time.

6. The AI-driven autonomous quantitative investment application technology system according to claim 1, characterized in that: The risk system in S6 includes the following steps: A module for generating real-time risk monitoring: This module collects real-time market data, assesses whether market price fluctuations meet expected prices, and determines whether profit and loss situations meet investment expectations. Risk assessment and analysis: Utilizing machine learning algorithms and techniques to process and analyze collected data, assess the current risk level of the investment portfolio, and identify potential risk factors; Set risk thresholds: Based on the user's risk preferences and investment goals, set reasonable risk thresholds as the standard for triggering risk control mechanisms; Develop risk control strategies: Once the risk exceeds the set threshold, the risk control module will be automatically triggered, generating corresponding risk control measures, such as increasing holdings, reducing holdings, or selling off, based on pre-set strategies and algorithms. Implement risk control measures: The system automatically adjusts the portfolio allocation based on the generated risk control measures to reduce the risk level.

7. The AI-driven autonomous quantitative investment application technology system according to claim 1, characterized in that: The risk control module uses logistic regression, a machine learning algorithm, to assess risk. This algorithm mines and analyzes large amounts of data to help the system more accurately identify potential risks and formulate corresponding risk control strategies. The formula is as follows: P(y=1|x) represents the probability that the output variable y is 1 given the input variable x. w is the weight vector, x is the input variable vector, b is the bias term, and e is the base of the natural logarithm. When setting a risk threshold, y=1 is interpreted as the occurrence of a risk event, i.e., the portfolio loss exceeds a certain threshold, while y=0 is interpreted as the risk event not occurring. By training the logistic regression algorithm, we can learn the optimal values ​​of the weight vector w and the bias term b, enabling the investment optimization model to accurately predict the probability of risk events. This helps us understand the model's predictive ability and adjust the model parameters accordingly to automatically adjust the portfolio allocation to reduce the risk level.

Images