Business data processing method and device, electronic equipment and computer program product

Abstract

The invention discloses a business data processing method and device, electronic equipment and a computer program product. Relates to the field of artificial intelligence, and the method comprises the steps: obtaining a group of business data, the data in the group of business data having an association relationship; inputting the group of service data into a decision feedback model to obtain a group of processed target service data, the decision feedback model being used for eliminating a non-linear relationship in the group of service data; and inputting the group of target business data into a data processing model to obtain a business data processing result, the data processing model being obtained by training N groups of training samples, each group of training samples comprising a group of historical business data and a historical business data processing result, and N being a positive integer. Through the method and the device, the problem of inaccurate service data processing result caused by a nonlinear relationship in the service data in the related technology is solved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence, and more specifically, to a method, device, electronic device, and computer program product for processing business data. Background Art

[0002] In the operation and management of financial markets, data accuracy and timeliness are the cornerstones of ensuring effective investment decisions and precise risk management. As financial markets continue to grow in globalization and complexity, massive amounts of real-time data are providing investors and financial institutions with unprecedented insights and opportunities. However, financial market data encompasses not only stock prices, trading volume, and market indices, but also complex economic and market sentiment factors. These factors make financial data inherently difficult to predict and highly uncertain. Financial data is susceptible to interference from unexpected factors, such as market fluctuations, trading errors, and policy changes. This random noise can arise from multiple sources, including errors in the data collection process, nonlinear market reactions, and the unpredictability of external events, significantly impacting data accuracy and availability.

[0003] Financial data analysis and forecasting in related technologies primarily rely on static models that rely on past data trends and patterns. This approach performs well when data is relatively stable and regular, but its forecasting accuracy and responsiveness are limited in rapidly changing market environments. Risk forecasting methods in related technologies rely on the experience and intuition of professional analysts, including reading market reports, analyzing economic indicators, and evaluating investment portfolios. While this approach offers strong interpretive power in complex situations, the subjectivity and latency of manual analysis limit its application in large-scale, real-time data processing. The primary drawback of relying on manual analysis is the time and workload. Analysts must spend considerable time researching various data sources, and are prone to omissions and errors when processing large amounts of data.

[0004] Currently, no effective solution has been proposed to the problem in related technologies that business data processing results are inaccurate due to nonlinear relationships in business data. Summary of the Invention

[0005] The main purpose of this application is to provide a business data processing method, device, electronic device and computer program product to solve the problem in related technologies that business data processing results are inaccurate due to nonlinear relationships in business data.

[0006] To achieve the above-mentioned objectives, according to one aspect of the present application, a method for processing business data is provided. The method comprises: obtaining a set of business data, wherein there is an association relationship between the data in the set of business data; inputting the set of business data into a decision feedback model to obtain a processed set of target business data, wherein the decision feedback model is used to eliminate nonlinear relationships in the set of business data; and inputting the set of target business data into a data processing model to obtain business data processing results, wherein the data processing model is trained using N sets of training samples, each set of training samples including a set of historical business data and a historical business data processing result, where N is a positive integer.

[0007] Optionally, the decision feedback model is obtained by: initializing the parameters of the decision feedback model, wherein the decision feedback model includes a forward filter, a reverse filter and a decision maker, and the parameters include at least the filter coefficients; obtaining M groups of historical business data, and determining the actual business data corresponding to each group of historical business data after eliminating nonlinear relationships and uncertainties, wherein M is a positive integer; determining each group of historical business data and the actual business data corresponding to the historical business data as a group of sample data to obtain M groups of sample data; iteratively training the decision feedback model through the M groups of sample data to obtain a trained decision feedback model, wherein each iterative training adjusts the parameters of the decision feedback model.

[0008] Optionally, the decision feedback model is iteratively trained by M groups of sample data to obtain the trained decision feedback model, including: for each iterative training, inputting a group of historical business data in each group of sample data into the forward filter to obtain initial predicted business data; inputting the initial predicted business data into the decision maker to obtain a decision result, and inputting the decision result into the reverse filter to obtain predicted business data; calculating the error value between the predicted business data and the actual business data of each group of sample data to obtain M error values, and calculating the average error value of the M error values; when the average error value of the iterative training of the target number of times is less than the error threshold, the model after the iterative training of the target number of times is determined as the trained decision feedback model.

[0009] Optionally, the data processing model is trained in the following manner: obtaining N groups of historical business data and determining the historical business data processing results corresponding to each group of historical business data; determining each group of historical business data and the historical business data processing results corresponding to the historical business data as a group of training samples to obtain N groups of training samples; training the machine learning model through the N groups of training samples to obtain the data processing model.

[0010] Optionally, after obtaining the business data processing result, the method further includes: judging whether the business data processing result meets the preset business data processing index; if the business data processing result does not meet the business data processing index, updating the decision feedback model.

[0011] Optionally, updating the decision feedback model includes: iteratively updating the decision feedback model to obtain an updated decision feedback model, wherein each iterative update adjusts the parameters of the decision feedback model; inputting a set of business data into the updated decision feedback model to obtain updated target business data; inputting the updated target business data into the data processing model to obtain an updated business data processing result; if the updated business data processing result does not meet the business data processing indicators, executing the step of iteratively updating the decision feedback model to obtain an updated decision feedback model; if the updated business data processing result meets the business data processing indicators, stopping updating the decision feedback model.

[0012] Optionally, before inputting a set of business data into the decision feedback model, the method further includes: preprocessing the set of business data to obtain a set of preprocessed business data, wherein the preprocessing includes at least one of the following: removing redundant data and empty data and normalization processing.

[0013] To achieve the above-mentioned purpose, according to another aspect of the present application, a device for processing business data is provided. The device includes: an acquisition unit for acquiring a set of business data, wherein there is an association relationship between the data in the set of business data; a first input unit for inputting the set of business data into a decision feedback model to obtain a processed set of target business data, wherein the decision feedback model is used to eliminate nonlinear relationships in the set of business data; and a second input unit for inputting the set of target business data into a data processing model to obtain a business data processing result, wherein the data processing model is trained by N sets of training samples, each set of training samples including a set of historical business data and a historical business data processing result, where N is a positive integer.

[0014] In an embodiment of the present application, a group of business data is obtained, wherein there is a correlation relationship between the data in the group of business data; the group of business data is input into a decision feedback model to obtain a processed group of target business data, wherein the decision feedback model is used to eliminate the nonlinear relationship in the group of business data; the group of target business data is input into a data processing model to obtain a business data processing result, wherein the data processing model is trained by N groups of training samples, each group of training samples includes a group of historical business data and the historical business data processing results, and N is a positive integer. The nonlinear relationship and uncertainty in the business data are eliminated through the decision feedback model, and then the denoised target business data is processed based on the data processing model, thereby achieving the purpose of removing noise in the financial data, thereby achieving the technical effect of improving the accuracy of the business data processing results, and thus solving the technical problem of inaccurate business data processing results due to the existence of nonlinear relationships in the business data. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The accompanying drawings, which constitute part of this application, are intended to provide a further understanding of this application. The exemplary embodiments and descriptions of this application are intended to explain this application and do not constitute an improper limitation on this application. In the accompanying drawings:

[0016] Figure 1 A hardware structure block diagram of a computer terminal (or mobile device) for implementing a method for processing business data is shown;

[0017] Figure 2 is a flowchart of a method for processing business data provided in an embodiment of the present application;

[0018] Figure 3 is a schematic diagram of a business data processing system provided according to an embodiment of the present application;

[0019] Figure 4 is a schematic diagram of a device for processing business data provided in an embodiment of the present application;

[0020] Figure 5 This is a structural block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0021] In order to enable those skilled in the art to better understand the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the 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 in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of this application.

[0022] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in a sequence other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0023] It should be noted that the collected information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for display, data for analysis, etc.) involved in this application are information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of the relevant data comply with relevant laws, regulations and standards, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation portals for users to choose to authorize or refuse. For example, an interface is set up between this system and relevant users or institutions to provide users with corresponding operation portals for users to choose to agree or refuse the automated decision-making results; if the user chooses to refuse, the expert decision-making process will be entered.

[0024] Example 1

[0025] According to an embodiment of the present application, an embodiment of a method for processing business data is also provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0026] The method embodiment provided in the first embodiment of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 1 The hardware structure block diagram of a computer terminal (or mobile device) for implementing a method for processing business data is shown. Figure 1 As shown, the computer terminal 10 (or mobile device) may include one or more (illustrated as 102a, 102b, ..., 102n in the figure) processors 102 (the processor 102 may include but is not limited to a processing device such as an MCU (Microcontroller Unit) or an FPGA (Field-Programmable Gate Array), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a USB (Universal Serial Bus) port (which may be included as one of the ports of a BUS (Business, bus)), a network interface, a power supply and / or a camera. It will be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 1 More or fewer components than shown, or with Figure 1 Different configurations shown.

[0027] It should be noted that the one or more processors 102 and / or other data processing circuits described above may generally be referred to herein as "data processing circuitry". The data processing circuitry may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuitry may be a single independent processing module, or may be incorporated in whole or in part into any of the other components of the computer terminal 10 (or mobile device). As described in the embodiments of the present application, the data processing circuitry serves as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).

[0028] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the business data processing method in the embodiment of the present application. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, realizes the above-mentioned business data processing method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include a memory remotely located relative to the processor 102, and these remote memories may be connected to the computer terminal 10 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0029] The transmission device 106 is configured to receive or transmit data via a network. A specific example of the aforementioned network may include a wireless network provided by the communications provider of the computer terminal 10. In one embodiment, the transmission device 106 includes a network interface controller (NIC), which can be connected to other network devices via a base station to enable communication with the Internet. In another embodiment, the transmission device 106 may be a radio frequency (RF) module, which is configured to communicate with the Internet wirelessly.

[0030] The display may be, for example, a touch screen liquid crystal display that enables a user to interact with a user interface of the computer terminal 10 (or mobile device).

[0031] In the above operating environment, this application provides a method for processing business data. Figure 2 is a flowchart of a method for processing business data according to an embodiment of the present application. Figure 2 As shown, the method includes:

[0032] Step S201: Acquire a group of business data, wherein there is an association relationship between the data in the group of business data.

[0033] In step S201, business-related data is first collected from sources such as the enterprise's databases, log files, transaction systems, and user interface interaction records. This data may include, but is not limited to, financial statements, sales data, customer information, supply chain information, marketing campaign data, website traffic, and user behavior data. Business data is preprocessed to remove invalid or redundant data points, address missing values, and ensure data quality. Data from various sources is integrated into a unified database or data warehouse. Data normalization or standardization allows business data of different scales or types to be compared and analyzed within the same model.

[0034] Step S202: input a set of service data into a decision feedback model to obtain a set of processed target service data, wherein the decision feedback model is used to eliminate nonlinear relationships in the set of service data.

[0035] In step S202, the acquired set of business data may contain nonlinear relationships and uncertain data, which may cause certain noise interference to the business risk prediction. Therefore, a decision feedback model is trained to remove the noise interference in the business data, thereby more accurately predicting business risks. A set of business data is input into the trained decision feedback model. The forward filter processes the real-time data, the decision maker makes a prediction, and the reverse filter dynamically adjusts based on the prediction results and historical decision results to eliminate the interference of nonlinear relationships. After processing, the decision feedback model outputs a set of target business data. This set of target business data has theoretically eliminated the influence of nonlinear relationships and uncertainty as much as possible, thereby improving the accuracy of the prediction.

[0036] Step S203: input a set of target business data into the data processing model to obtain business data processing results, wherein the data processing model is trained by N sets of training samples, each set of training samples includes a set of historical business data and historical business data processing results, and N is a positive integer.

[0037] In step S203, historical business data and historical business data processing results are collected, ensuring that each set of historical business data is paired with the corresponding historical business data processing results, forming N sets of training samples. The machine learning model is trained using these N sets of training samples. A set of target business data is input into the trained data processing model. Based on its learned patterns and parameters, the model outputs business data processing results for the target set of business data.

[0038] The business data processing method provided in the embodiment of the present application obtains a group of business data, wherein there is an association relationship between the data in the group of business data; inputs the group of business data into a decision feedback model to obtain a processed group of target business data, wherein the decision feedback model is used to eliminate the nonlinear relationship in the group of business data; inputs the group of target business data into a data processing model to obtain a business data processing result, wherein the data processing model is trained by N groups of training samples, each group of training samples includes a group of historical business data and the historical business data processing results, N is a positive integer, eliminates the nonlinear relationship and uncertainty in the business data through the decision feedback model, and then processes the denoised target business data based on the data processing model, thereby achieving the purpose of removing noise in the financial data, thereby realizing the technical effect of improving the accuracy of the business data processing results, and thus solving the technical problem of inaccurate business data processing results due to the existence of nonlinear relationships in the business data.

[0039] In order to eliminate the nonlinear relationship and uncertainty in the business data, it is necessary to train a decision feedback model. Optionally, in the business data processing method provided in the embodiment of the present application, the decision feedback model is obtained by: initializing the parameters of the decision feedback model, wherein the decision feedback model includes a forward filter, a reverse filter and a decision maker, and the parameters include at least the filter coefficients; obtaining M groups of historical business data, and determining the actual business data corresponding to each group of historical business data after eliminating the nonlinear relationship and uncertainty, wherein M is a positive integer; determining each group of historical business data and the actual business data corresponding to the historical business data as a group of sample data to obtain M groups of sample data; iteratively training the decision feedback model through M groups of sample data to obtain a trained decision feedback model, wherein each iterative training adjusts the parameters of the decision feedback model.

[0040] In some embodiments, the decision feedback model may include a forward filter, a reverse filter, and a decision maker, each filter having its own specific number of taps, i.e., the number of parameters. The forward filter is used to process the input data to reduce inter-symbol interference (ISI) and noise, and the reverse filter is used to provide feedback based on the output of the decision maker to eliminate the influence of decision errors. During initialization, preliminary filter coefficients need to be set for both filters. The decision maker is responsible for making decisions based on the output of the forward filter, and its parameters are related to the decision threshold of the model. In some cases, the parameters of the decision maker can be estimated directly from the statistical properties of the data, such as the mean and standard deviation.

[0041] Collect M sets of historical business data, where M is a positive integer. Remove noise, outliers, and missing values, and perform data normalization to ensure the quality and consistency of the dataset. Through preprocessing and feature engineering, determine which data points in each set of historical business data are affected by nonlinear relationships and uncertainty. Identify and eliminate nonlinear relationships through data dimensionality reduction, feature selection, and techniques such as time series analysis and Granger causality testing. The processed data is closer to the actual state of the business and can serve as the "target" or "label" for model training. Pair the historical business data with its corresponding actual business data after eliminating nonlinear relationships and uncertainties to form M sets of sample data.

[0042] Defines the total number of training iterations. During each training iteration, the model learns based on the current parameter settings and sample data, adjusting the filter coefficients and decision maker parameters based on prediction error. A validation dataset, not used in training, is used to evaluate the model's generalization ability, ensuring that the model not only fits the training data but also maintains good predictive performance on new data. The trained and validated model can be used to process real-time business data, eliminating the impact of nonlinear relationships and improving the accuracy of risk predictions or business decisions.

[0043] This embodiment effectively eliminates nonlinear relationships and uncertainties in business data by training a decision feedback model, thereby improving data accuracy and reliability of business data processing.

[0044] After determining the sample data, a decision feedback model is trained based on M groups of sample data. Optionally, in the business data processing method provided in an embodiment of the present application, the decision feedback model is iteratively trained using M groups of sample data to obtain a trained decision feedback model, including: for each iterative training, inputting a group of historical business data in each group of sample data into a forward filter to obtain initial predicted business data; inputting the initial predicted business data into a decision maker to obtain a decision result, and inputting the decision result into a reverse filter to obtain predicted business data; calculating the error value between the predicted business data and the actual business data of each group of sample data to obtain M error values, and calculating the average error value of the M error values; when the average error value of the iterative training of the target number of times is less than the error threshold, determining the model after the iterative training of the target number of times as the trained decision feedback model.

[0045] In some embodiments, a set of historical business data is input into a forward filter. The function of the forward filter is to adjust its filter coefficients based on the historical data to reduce the impact of inter-code interference and noise and generate an initial predicted business data. The coefficients of the forward filter are fine-tuned according to the training objectives in each iteration in order to more accurately capture the dynamic characteristics of the business data. The initial predicted business data output by the forward filter is fed into the decision maker, which generates a decision result based on a preset decision rule or threshold. In financial business, it can be a classification of market trends (such as rising, falling or stable). The output of the decision maker (decision result) is fed back to the reverse filter. The reverse filter adjusts its filter coefficients according to the decision result to generate more accurate predicted business data. The coefficients of the reverse filter are also adjusted in each iteration to reduce the impact of decision errors on subsequent predictions.

[0046] Compare the final predicted business data output by the inverse filter with the actual business data, and calculate the error value between the two. For M groups of sample data, M error values ​​will be obtained. Average the M error values ​​to obtain the average error value. The average error value reflects the overall performance of the model's prediction ability for M groups of sample data under the current parameter settings. Set an error threshold, which is based on business needs and model performance requirements. When the average error value after the target number of iterative training is less than this error threshold, it means that the model has achieved a sufficiently low prediction error and meets the preset optimization goal. During the iterative training process, if the average error value is less than the error threshold, the model parameters at this time can be fixed, and the model can be determined as a trained decision feedback model.

[0047] This embodiment iteratively trains the decision feedback model through M groups of sample data until the average error value of the target number of iterative training is less than the error threshold. This can improve the prediction performance of the decision feedback model and more accurately remove noise interference when processing complex and nonlinear financial data.

[0048] In order to process business data, it is necessary to train a data processing model. Optionally, in the business data processing method provided in the embodiment of the present application, the data processing model is trained in the following manner: obtaining N groups of historical business data and determining the historical business data processing results corresponding to each group of historical business data; determining each group of historical business data and the historical business data processing results corresponding to the historical business data as a group of training samples to obtain N groups of training samples; training a machine learning model through the N groups of training samples to obtain a data processing model.

[0049] In some embodiments, N groups of historical data related to a specific business are collected. These data may include transaction records, customer behavior, market indicators, financial reports, etc. For each group of historical business data, the historical business data processing results at that time are collected or calculated, and each group of historical business data is paired with its corresponding historical business data processing results to form N groups of training samples. According to the characteristics of the business data (such as linear, nonlinear, time series, etc.) and the specific needs of data processing, a suitable machine learning model is selected. It may include but is not limited to logistic regression, support vector machine, decision tree, random forest, neural network, etc. The selected machine learning model is trained using N groups of training samples. During the training process, the model will learn the relationship between the data features and the business data processing results, and adjust its internal parameters to minimize the prediction error. During the training process, a loss function (such as mean square error or cross entropy) is defined to measure the gap between the model's prediction results and the actual historical risk prediction results. The model parameters are adjusted through an optimization algorithm (such as gradient descent) to minimize this loss function.

[0050] This embodiment learns business data processing patterns from N sets of historical business data. The trained data processing model can process business data more accurately, providing financial institutions with powerful decision support tools in a complex and changing market environment, while improving the degree of automation of predictions and resource utilization efficiency.

[0051] In order to ensure the accuracy of business data processing, the decision feedback model is updated when the business data processing result is erroneous. Optionally, in the business data processing method provided in the embodiment of the present application, after obtaining the business data processing result, the method also includes: judging whether the business data processing result meets the preset business data processing indicators; if the business data processing result does not meet the business data processing indicators, updating the decision feedback model.

[0052] In some embodiments, business data processing indicators for the target business data are determined based on actual business practices and expert judgment. The business data processing results of the model are compared with the business data processing indicators to determine whether they meet the preset business data processing indicators. If they do not meet the standards, this indicates an error in the business data processing and requires an update to the decision feedback model.

[0053] Retrain the decision feedback model using additional business data or adjusting model parameters, such as increasing the training sample size, adjusting filter coefficients, or updating the decision maker logic. Repeat the model training and testing process until the business data processing results of the model meet the business data processing indicators.

[0054] This embodiment can significantly enhance the overall performance and adaptability of the data processing model through dynamic updating of the decision feedback model.

[0055] If the prediction accuracy is lower than the accuracy threshold, the decision feedback model needs to be updated. Optionally, in the business data processing method provided in the embodiment of the present application, updating the decision feedback model includes: iteratively updating the decision feedback model to obtain an updated decision feedback model, wherein each iterative update adjusts the parameters of the decision feedback model; inputting a set of business data into the updated decision feedback model to obtain updated target business data; inputting the updated target business data into the data processing model to obtain an updated business data processing result; if the updated business data processing result does not meet the business data processing index, performing the step of iteratively updating the decision feedback model to obtain an updated decision feedback model; if the updated business data processing result meets the business data processing index, stopping updating the decision feedback model.

[0056] In some embodiments, the decision feedback model is iteratively trained using historical business data and its corresponding business data processing results as a training set. A set of business data is input into the iteratively updated decision feedback model, and updated target business data is obtained through model processing. The updated target business data processed by the decision feedback model is input into the data processing model to obtain an updated business data processing result. If the updated business data processing result does not meet the business data processing index, it means that the decision feedback model needs to be further iteratively updated. If the updated risk prediction result meets the business data processing index, or the error is within an acceptable range, it indicates that the model has achieved the ideal prediction performance, and the update process of the decision feedback model can be stopped.

[0057] This embodiment dynamically optimizes the decision feedback model to more accurately process business data, reduce the impact of nonlinear relationships and uncertainty on risk prediction, and ultimately improve the accuracy of the data processing model.

[0058] Improve model processing efficiency through preprocessing. Optionally, in the business data processing method provided in an embodiment of the present application, before inputting a set of business data into the decision feedback model, the method further includes: preprocessing a set of business data to obtain a set of preprocessed business data, wherein the preprocessing includes at least one of the following: removing redundant data and empty data and normalization processing.

[0059] In some embodiments, redundant data refers to repeated or redundant information in the data set, which does not provide additional value for modeling, but may increase the computational burden or cause the model to overfit. Through data cleaning, redundant data entries are identified and removed. For example, if there are multiple records in the database pointing to different aspects of the same transaction, these records can be merged to retain only the most relevant or complete information. Check whether there are missing values ​​or empty data in the data set. Empty data may be caused by errors in the data collection process or the inability to obtain certain information. Methods for handling empty data include deleting records containing empty values, filling empty values ​​(using the mean, median, mode or other predicted values) or using data interpolation techniques. Normalization is a data transformation technique used to unify the data scale to a specific range, such as the interval from 0 to 1.

[0060] Through the above preprocessing steps, the quality of business data can be improved, ensuring that the data input into the decision feedback model and risk prediction model is clean and consistent, thereby helping the model learn data patterns more accurately, improving prediction accuracy and model robustness.

[0061] According to another embodiment of the present application, a business data processing system is also provided. Figure 3 is a schematic diagram of a business data processing system provided according to an embodiment of the present application, such as Figure 3 As shown, the system includes: a data acquisition module, a data preprocessing module, and a decision feedback module.

[0062] The data acquisition module can accept financial data related to stock prices, trading volumes, and market indices as input. Secondly, the data preprocessing module removes redundant and null data from the real data and, through normalization, linearly scales the data to the (0, 1) range. Next, the decision feedback model is trained using historical financial data. Based on the deviation between the model output and the actual results, the model parameters are adjusted to improve the accuracy of the model's predictions, enabling it to distinguish between normal and abnormal data. Finally, the trained model is used to classify and predict new financial data, outputting accurate decision results.

[0063] Since the currently input financial data is affected by the tail interference of the previously input financial data and the leading interference of the subsequent financial data, after the current financial data is detected and determined, the interference of the current financial data on the subsequent financial data can be eliminated before checking the subsequent financial data. k , the output of the feedforward filter is: Among them, (-k1<<k<<0), f n Represents the equivalent pre-tap coefficient, and the input to the decision maker is: In the above formula, the first f j(-k1≤j≤0) represents the tap coefficient of the equivalent prefilter. The second f j (1≤j≤k2) represents the tap coefficient of the equivalent feedback filter. According to the minimum mean square error principle, assuming that the judgment of the previous signal is correct, the minimum mean square error is defined as: J(k1,k2)=E|I k -I ′ k | 2 At this time, the forward filter coefficient can be obtained by minimizing J(k1,k2), and the feedback filter coefficient can be obtained by the pre-filter coefficient. By adjusting the filter coefficient, the final correct judgment result is obtained to achieve the prediction of financial data.

[0064] The business data processing system of this embodiment establishes a financial data prediction device by modifying filter coefficients. By continuously learning historical data to automatically adjust the filter parameters, nonlinear relationships and uncertainties in business data are eliminated.

[0065] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0066] Example 2

[0067] The present application also provides a business data processing device. It should be noted that the business data processing device of the present application can be used to execute the business data processing method provided in the present application. The business data processing device provided in the present application is introduced below.

[0068] According to an embodiment of the present application, a device for implementing the above-mentioned method for processing business data is also provided. Figure 4 is a schematic diagram of a device for processing business data according to an embodiment of the present application, such as Figure 4 As shown, the device includes:

[0069] The acquiring unit 401 is configured to acquire a set of business data, wherein there is an association relationship between the data in the set of business data.

[0070] The first input unit 402 is configured to input the set of service data into a decision feedback model to obtain a processed set of target service data, wherein the decision feedback model is configured to eliminate nonlinear relationships in the set of service data.

[0071] The second input unit 403 is used to input the set of target business data into the data processing model to obtain a business data processing result, wherein the data processing model is trained by N sets of training samples, each set of training samples includes a set of historical business data and historical business data processing results, and N is a positive integer.

[0072] The business data processing device provided in the embodiment of the present application obtains a group of business data through an acquisition unit 401, wherein there is a correlation relationship between the data in the group of business data; a first input unit 402 inputs the group of business data into a decision feedback model to obtain a processed group of target business data, wherein the decision feedback model is used to eliminate the nonlinear relationship in the group of business data; a second input unit 403 inputs the group of target business data into a data processing model to obtain a business data processing result, wherein the data processing model is trained by N groups of training samples, each group of training samples includes a group of historical business data and the historical business data processing result, N is a positive integer, and the nonlinear relationship and uncertainty in the business data are eliminated through the decision feedback model, and then the denoised target business data is processed based on the data processing model, thereby achieving the purpose of removing noise in the financial data, thereby realizing the technical effect of improving the accuracy of the business data processing results, and thus solving the technical problem of inaccurate business data processing results due to the existence of nonlinear relationships in the business data.

[0073] Optionally, in the business data processing device provided in the embodiment of the present application, the device also includes: an initialization unit, used to initialize the parameters of the decision feedback model, wherein the decision feedback model includes a forward filter, a reverse filter and a decision maker, and the parameters include at least filter coefficients; a first determination unit, used to obtain M groups of historical business data, and determine the actual business data corresponding to each group of historical business data after eliminating nonlinear relationships and uncertainties, wherein M is a positive integer; a second determination unit, used to determine each group of historical business data and the actual business data corresponding to the historical business data as a group of sample data, to obtain M groups of sample data; a first training unit, used to iteratively train the decision feedback model through M groups of sample data to obtain a trained decision feedback model, wherein each iterative training adjusts the parameters of the decision feedback model.

[0074] Optionally, in the business data processing device provided in the embodiment of the present application, the first training unit includes: a first input module, which is used to input a group of historical business data in each group of sample data into a forward filter for each iterative training to obtain initial predicted business data; a second input module, which is used to input the initial predicted business data into a decision maker to obtain a decision result, and input the decision result into a reverse filter to obtain predicted business data; a calculation module, which is used to calculate the error value between the predicted business data and the actual business data of each group of sample data, obtain M error values, and calculate the average error value of the M error values; a first determination module, which is used to determine the model after the target number of iterative training as the trained decision feedback model when the average error value of the target number of iterative training is less than the error threshold.

[0075] Optionally, in the business data processing device provided in the embodiment of the present application, the device also includes: a third determination unit, used to obtain N groups of historical business data and determine the historical business data processing results corresponding to each group of historical business data; a fourth determination unit, used to determine each group of historical business data and the historical business data processing results corresponding to the historical business data as a group of training samples to obtain N groups of training samples; a second training unit, used to train a machine learning model through N groups of training samples to obtain a data processing model.

[0076] Optionally, in the business data processing device provided in the embodiment of the present application, the device also includes: a judgment unit, used to judge whether the business data processing result meets the preset business data processing indicators; and an update unit, used to update the decision feedback model when the business data processing result does not meet the business data processing indicators.

[0077] Optionally, in the business data processing device provided in the embodiment of the present application, the update unit includes: an update module, which is used to iteratively update the decision feedback model to obtain an updated decision feedback model, wherein each iterative update adjusts the parameters of the decision feedback model; a third input module, which is used to input a set of business data into the updated decision feedback model to obtain updated target business data; a fourth input module, which is used to input the updated target business data into the data processing model to obtain an updated business data processing result; an execution module, which is used to iteratively update the decision feedback model to obtain an updated decision feedback model when the updated business data processing result does not meet the business data processing indicators; and a stop module, which is used to stop updating the decision feedback model when the updated business data processing result meets the business data processing indicators.

[0078] Optionally, in the business data processing device provided in an embodiment of the present application, the device also includes: a preprocessing unit, used to preprocess a group of business data to obtain a preprocessed group of business data, wherein the preprocessing includes at least one of the following: removing redundant data and empty data and normalization processing.

[0079] It should be noted that the acquisition unit 401, the first input unit 402, and the second input unit 403 correspond to steps S201 to S203 in Example 1. The three units and the corresponding steps implement the same examples and application scenarios, but are not limited to the contents disclosed in the above-mentioned Example 1. It should be noted that the above-mentioned modules or units can be hardware components or software components stored in a memory (e.g., memory 104) and processed by one or more processors (e.g., processors 102a, 102b, ..., 102n). The above-mentioned modules or units can also be part of a device and can be run in the computer terminal 10 provided in Example 1.

[0080] Example 3

[0081] An embodiment of the present application may provide an electronic device, Figure 5 This is a structural block diagram of an electronic device according to an embodiment of the present application. Figure 5 As shown, the electronic device may include: one or more ( Figure 5 Only one is shown) processor 502, memory 504, storage controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module and display.

[0082] Among them, the memory can be used to store software programs and modules, such as program instructions / modules corresponding to the methods and devices in the embodiments of the present application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, that is, implementing the above-mentioned method. The memory may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include a memory remotely arranged relative to the processor, and these remote memories may be connected to the terminal via a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network and a combination thereof.

[0083] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: obtain a set of business data, wherein there is a correlation relationship between the data in the set of business data; input the set of business data into the decision feedback model to obtain a processed set of target business data, wherein the decision feedback model is used to eliminate the nonlinear relationship in the set of business data; input the set of target business data into the data processing model to obtain the business data processing results, wherein the data processing model is trained by N sets of training samples, each set of training samples includes a set of historical business data and the historical business data processing results, and N is a positive integer.

[0084] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: initialize the parameters of the decision feedback model, wherein the decision feedback model includes a forward filter, a reverse filter and a decision maker, and the parameters include at least the filter coefficients; obtain M groups of historical business data, and determine the actual business data corresponding to each group of historical business data after eliminating nonlinear relationships and uncertainties, wherein M is a positive integer; determine each group of historical business data and the actual business data corresponding to the historical business data as a group of sample data to obtain M groups of sample data; iteratively train the decision feedback model through M groups of sample data to obtain a trained decision feedback model, wherein each iterative training adjusts the parameters of the decision feedback model.

[0085] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: for each iterative training, a set of historical business data in each set of sample data is input into the forward filter to obtain initial predicted business data; the initial predicted business data is input into the decision maker to obtain a decision result, and the decision result is input into the reverse filter to obtain predicted business data; the error value between the predicted business data and the actual business data of each set of sample data is calculated to obtain M error values, and the average error value of the M error values ​​is calculated; when the average error value of the iterative training of the target number of times is less than the error threshold, the model after the iterative training of the target number of times is determined as the trained decision feedback model.

[0086] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: obtain N groups of historical business data and determine the historical business data processing results corresponding to each group of historical business data; determine each group of historical business data and the historical business data processing results corresponding to the historical business data as a group of training samples to obtain N groups of training samples; train the machine learning model through the N groups of training samples to obtain a data processing model.

[0087] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: determine whether the business data processing results meet the preset business data processing indicators; if the business data processing results do not meet the business data processing indicators, update the decision feedback model.

[0088] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: iteratively update the decision feedback model to obtain an updated decision feedback model, wherein each iterative update adjusts the parameters of the decision feedback model; input a set of business data into the updated decision feedback model to obtain updated target business data; input the updated target business data into the data processing model to obtain an updated business data processing result; if the updated business data processing result does not meet the business data processing index, perform the step of iteratively updating the decision feedback model to obtain an updated decision feedback model; if the updated business data processing result meets the business data processing index, stop updating the decision feedback model.

[0089] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: preprocessing a set of business data to obtain a preprocessed set of business data, wherein the preprocessing includes at least one of the following: removing redundant data and empty data and normalization processing.

[0090] An embodiment of the present application provides a method for obtaining a set of business data, wherein there is an association relationship between the data in the set of business data; inputting the set of business data into a decision feedback model to obtain a processed set of target business data, wherein the decision feedback model is used to eliminate nonlinear relationships in the set of business data; inputting the set of target business data into a data processing model to obtain business data processing results, wherein the data processing model is trained by N sets of training samples, each set of training samples includes a set of historical business data and historical business data processing results, and N is a positive integer. The nonlinear relationship and uncertainty in the business data are eliminated by the decision feedback model, and then the denoised target business data is processed based on the data processing model, thereby achieving the purpose of removing noise in the financial data, thereby achieving the technical effect of improving the accuracy of the business data processing results, and thus solving the technical problem of inaccurate business data processing results due to the presence of nonlinear relationships in the business data.

[0091] It can be understood by those skilled in the art that Figure 5 The structure shown is for illustration only, and the electronic device may also be a terminal device such as a smart phone, a tablet computer, a PDA, a mobile Internet device (MID), or a PAD. Figure 5It does not limit the structure of the above electronic device. For example, the electronic device may also include Figure 5 More or fewer components (such as network interfaces, display devices, etc.) shown in, or with Figure 5 Different configurations shown.

[0092] A person skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing the hardware related to the terminal device through a program, and the program can be stored in a computer-readable storage medium, which may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0093] Example 4

[0094] The embodiment of the present application further provides a storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the method for processing business data provided in the first embodiment.

[0095] Optionally, in this embodiment, the storage medium may be located in any computer terminal in a computer terminal group in a computer network, or in any mobile terminal in a mobile terminal group.

[0096] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing the steps of the method for processing business data.

[0097] The serial numbers of the above embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0098] In the above embodiments of the present application, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0099] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0100] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0101] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0102] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for enabling a computer device (which can be a personal computer, a server or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk.

[0103] The above is only a preferred embodiment of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for processing business data, characterized in that: include: Acquire a set of business data, wherein there is an association relationship between data in the set of business data; Inputting the set of service data into a decision feedback model to obtain a processed set of target service data, wherein the decision feedback model is used to eliminate nonlinear relationships in the set of service data; The set of target business data is input into a data processing model to obtain a business data processing result, wherein the data processing model is trained by N sets of training samples, each set of training samples includes a set of historical business data and historical business data processing results, and N is a positive integer.

2. The method according to claim 1, characterized in that The decision feedback model is obtained in the following way: Initializing parameters of a decision feedback model, wherein the decision feedback model includes a forward filter, a reverse filter, and a decision maker, and the parameters include at least filter coefficients; Obtain M groups of historical business data, and determine the actual business data corresponding to each group of historical business data after eliminating nonlinear relationships and uncertainties, where M is a positive integer; Determine each set of historical business data and the actual business data corresponding to the historical business data as a set of sample data, to obtain M sets of sample data; The decision feedback model is iteratively trained using the M groups of sample data to obtain a trained decision feedback model, wherein parameters of the decision feedback model are adjusted during each iterative training.

3. The method according to claim 2, characterized in that The decision feedback model is iteratively trained by the M groups of sample data to obtain a trained decision feedback model including: For each iterative training, a set of historical business data in each set of sample data is input into the forward filter to obtain initial predicted business data; Inputting the initial predicted business data into the decision maker to obtain a decision result, and inputting the decision result into the inverse filter to obtain predicted business data; Calculate the error value between the predicted business data and the actual business data of each set of sample data to obtain M error values, and calculate the average error value of the M error values; When the average error value of the iterative training of the target number of times is less than the error threshold, the model after the iterative training of the target number of times is determined as the trained decision feedback model.

4. The method according to claim 1, wherein The data processing model is trained in the following way: Obtain N groups of historical business data and determine the historical business data processing results corresponding to each group of historical business data; Determine each set of historical business data and the historical business data processing results corresponding to the historical business data as a set of training samples, to obtain N sets of training samples; The machine learning model is trained using the N groups of training samples to obtain the data processing model.

5. The method according to claim 1, wherein After obtaining the business data processing result, the method further includes: Determine whether the business data processing result meets the preset business data processing indicators; When the service data processing result does not meet the service data processing indicator, the decision feedback model is updated.

6. The method according to claim 5, characterized in that Updating the decision feedback model includes: Iteratively updating the decision feedback model to obtain an updated decision feedback model, wherein each iterative update adjusts parameters of the decision feedback model; Inputting the set of service data into the updated decision feedback model to obtain updated target service data; Inputting the updated target business data into the data processing model to obtain an updated business data processing result; If the updated service data processing result does not meet the service data processing indicator, performing the step of iteratively updating the decision feedback model to obtain an updated decision feedback model; When the updated service data processing result meets the service data processing indicator, the updating of the decision feedback model is stopped.

7. The method according to claim 1, characterized in that Before inputting the set of service data into the decision feedback model, the method further includes: The set of business data is preprocessed to obtain a set of preprocessed business data, wherein the preprocessing includes at least one of the following: removing redundant data and empty data and normalization processing.

8. A device for processing business data, characterized in that: include: An acquiring unit, configured to acquire a set of business data, wherein there is an association relationship between the data in the set of business data; a first input unit, configured to input the set of service data into a decision feedback model to obtain a processed set of target service data, wherein the decision feedback model is configured to eliminate nonlinear relationships in the set of service data; The second input unit is used to input the set of target business data into the data processing model to obtain the business data processing results, wherein the data processing model is trained by N sets of training samples, each set of training samples includes a set of historical business data and historical business data processing results, and N is a positive integer.

9. An electronic device, characterized in that: include: a memory storing an executable program; A processor is used to run the program, wherein when the program is run, the business data processing method described in any one of claims 1 to 7 is executed.

10. A computer program product comprising computer instructions, characterized in that When the computer instructions are executed by a processor, the steps of the business data processing method described in any one of claims 1 to 7 are implemented.

Images