Data prediction method and device based on artificial intelligence, computer equipment and medium
Through the data prediction method based on artificial intelligence, the gradient-enhancing decision-making forest model is used to predict medical expenses, which solves the prediction efficiency and accuracy of medical project expenses under traditional artificial methods, and achieves fast and accurate medical expense prediction.
Patent Information
- Application Number
- CN202411173233.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-26
- Publication Date
- 2024-12-06
AI Technical Summary
In the prior art, the prediction method of medical project costs depends on manual methods, which is time-consuming and labor-intensive, inefficient and susceptible to human factors, resulting in inaccurate prediction results.
Using an artificial intelligence-based data prediction method, by obtaining the user's initial electronic medical record information, preprocessing and information expansion, we call the gradient-enhancing decision-making forest model built on medical record sample data to predict medical expenses.
It improves the processing efficiency and accuracy of medical cost prediction, ensures the accuracy of generated prediction values, reduces human interference, and improves the reliability of prediction results.
Smart Images

Figure CN119090070A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of artificial intelligence development technology and the field of digital medicine, and in particular to artificial intelligence-based data prediction methods, devices, computer equipment and storage media. Background Art
[0002] In the current insurance industry, with the continuous innovation of insurance products and the increasing diversification of market demand, insurance with illness has become a trend that cannot be ignored. This change has brought unprecedented risk assessment and pricing challenges to insurance companies. Specifically, insurance companies need to more accurately predict the risk level of insured persons with illnesses, including their possible future medical expenses, such as medicine expenses, hospitalization expenses, etc., and dynamically adjust insurance premiums accordingly to ensure a reasonable match between risk and premium.
[0003] At the same time, within medical institutions, especially large medical service providers such as hospitals, each department arranges corresponding diagnosis and treatment items according to the patient's specific condition and issues corresponding medical expense receipts. This process involves a large amount of medical project cost data, and its accuracy and rationality directly affect the patient's economic burden and the rational allocation of medical resources. Therefore, a quality control department is set up within the medical institution to supervise the rationality of the charging items of each department, prevent drug abuse, unnecessary examinations and treatment items, and ensure the fairness and transparency of medical expenses.
[0004] However, under the traditional technical framework, the quality control department often relies on manual methods to predict and supervise the cost of medical projects. This requires relevant personnel to frequently shuttle between departments to manually collect, organize and analyze the charging data of various medical projects, and then make cost predictions and rationality assessments based on these data. This method is not only time-consuming and labor-intensive, inefficient, but also easily interfered by human factors, resulting in inaccurate prediction results, which in turn affects the effectiveness and timeliness of quality control work. Summary of the invention
[0005] The purpose of the embodiments of the present application is to propose a data prediction method, device, computer equipment and storage medium based on artificial intelligence to solve the technical problem that the existing technology often relies on manual methods to predict the cost of medical projects, which is time-consuming, labor-intensive, inefficient, and easily affected by human factors, resulting in inaccurate prediction results.
[0006] In order to solve the above technical problems, the present application embodiment provides a data prediction method based on artificial intelligence, which adopts the following technical solutions:
[0007] Obtain the user's initial electronic medical record information;
[0008] Preprocessing the initial electronic medical record information to obtain corresponding electronic medical record information;
[0009] Extracting the user's personal information from the electronic medical record information, and obtaining the user's province information from the personal information;
[0010] Filtering target characteristic data matching the province information from the preset province characteristic data;
[0011] Performing information expansion processing on the electronic medical record information based on the target feature data to obtain corresponding target electronic medical record information;
[0012] Calling a pre-built cost prediction model; wherein the cost prediction model is generated by training a preset gradient boosting decision forest model based on medical record sample data, and the gradient boosting decision forest model is a model of gradient boosting decision tree combined with random forest;
[0013] The target electronic medical record information is predicted based on the cost prediction model to generate a medical cost prediction value corresponding to the target electronic medical record information.
[0014] Furthermore, the step of selecting target characteristic data matching the province information from the preset province characteristic data specifically includes:
[0015] Get preset public statistics;
[0016] Extracting the regional characteristic data and medical characteristic data of each province from the public statistical data to obtain the provincial characteristic data;
[0017] Filter out the designated regional characteristic data matching the provincial information from the regional characteristic data of each province;
[0018] Filter out designated medical characteristic data matching the provincial information from the medical characteristic data of each of the provinces;
[0019] The target characteristic data is constructed based on the designated regional characteristic data and the designated medical characteristic data.
[0020] Furthermore, the step of performing information expansion processing on the electronic medical record information based on the target feature data to obtain the corresponding target electronic medical record information specifically includes:
[0021] Call the preset data analysis tool;
[0022] Using a merge operation in the data analysis tool to merge the electronic medical record information with the target feature data to obtain corresponding merged data;
[0023] The combined data is used as the target feature data.
[0024] Furthermore, the step of preprocessing the initial electronic medical record information to obtain corresponding electronic medical record information specifically includes:
[0025] Performing data cleaning on the initial electronic medical record information to obtain corresponding first electronic medical record information;
[0026] Obtain the preset classification coding strategy;
[0027] Encoding the first electronic medical record information based on the classification coding strategy to obtain corresponding second electronic medical record information;
[0028] The second electronic medical record information is used as the electronic medical record information.
[0029] Furthermore, before the step of calling the pre-built cost prediction model, the method further includes:
[0030] Get pre-built sample medical record data;
[0031] Creating a random forest model, and using the medical record sample data to train the random forest model to obtain a corresponding target random forest model;
[0032] Creating a gradient boosting decision forest model and using the target random forest model as an initial prediction estimator for the gradient boosting decision forest model;
[0033] Determine a target loss function and a target hyperparameter tuning method corresponding to the gradient boosting decision forest model;
[0034] Based on the target loss function and the target hyperparameter tuning method, the gradient boosting decision forest model is trained with ten-fold cross validation using the medical record sample data to obtain a specified gradient boosting decision forest model that meets the preset model building conditions;
[0035] The specified gradient boosting decision forest model is used as the cost prediction model.
[0036] Further, after the step of using the designated gradient boosting decision forest model as the cost prediction model, the method further includes:
[0037] Obtaining a specified storage strategy corresponding to the cost prediction model;
[0038] Determining a designated storage medium corresponding to the cost prediction model;
[0039] The cost prediction model is stored based on the specified storage strategy and the specified storage medium.
[0040] Furthermore, after the step of performing prediction processing on the target electronic medical record information based on the cost prediction model to generate a medical cost prediction value corresponding to the target electronic medical record information, the method further includes:
[0041] Obtaining communication information of the user based on the personal information;
[0042] Obtaining explanation information corresponding to the predicted value of medical expenses;
[0043] Generate a corresponding medical expense prediction report based on the medical expense prediction value and the explanation information;
[0044] Based on the communication information, the medical expense prediction report is sent to the user.
[0045] In order to solve the above technical problems, the embodiment of the present application further provides a data prediction device based on artificial intelligence, which adopts the following technical solution:
[0046] The first acquisition module is used to acquire the user's initial electronic medical record information;
[0047] A preprocessing module, used to preprocess the initial electronic medical record information to obtain corresponding electronic medical record information;
[0048] An extraction module, used to extract the user's personal information from the electronic medical record information, and obtain the user's province information from the personal information;
[0049] A screening module, used to screen out target characteristic data matching the province information from preset province characteristic data;
[0050] An expansion module, used to perform information expansion processing on the electronic medical record information based on the target feature data to obtain corresponding target electronic medical record information;
[0051] A calling module, used to call a pre-built cost prediction model; wherein the cost prediction model is generated by training a preset gradient boosting decision forest model based on medical record sample data, and the gradient boosting decision forest model is a model of a gradient boosting decision tree combined with a random forest;
[0052] The prediction module is used to perform prediction processing on the target electronic medical record information based on the cost prediction model to generate a medical cost prediction value corresponding to the target electronic medical record information.
[0053] In order to solve the above technical problems, the embodiment of the present application further provides a computer device, which adopts the following technical solution:
[0054] Obtain the user's initial electronic medical record information;
[0055] Preprocessing the initial electronic medical record information to obtain corresponding electronic medical record information;
[0056] Extracting the user's personal information from the electronic medical record information, and obtaining the user's province information from the personal information;
[0057] Filtering target characteristic data matching the province information from the preset province characteristic data;
[0058] Performing information expansion processing on the electronic medical record information based on the target feature data to obtain corresponding target electronic medical record information;
[0059] Calling a pre-built cost prediction model; wherein the cost prediction model is generated by training a preset gradient boosting decision forest model based on medical record sample data, and the gradient boosting decision forest model is a model of gradient boosting decision tree combined with random forest;
[0060] The target electronic medical record information is predicted based on the cost prediction model to generate a medical cost prediction value corresponding to the target electronic medical record information.
[0061] In order to solve the above technical problems, the embodiment of the present application further provides a computer-readable storage medium, which adopts the following technical solution:
[0062] Obtain the user's initial electronic medical record information;
[0063] Preprocessing the initial electronic medical record information to obtain corresponding electronic medical record information;
[0064] Extracting the user's personal information from the electronic medical record information, and obtaining the user's province information from the personal information;
[0065] Filtering target characteristic data matching the province information from the preset province characteristic data;
[0066] Performing information expansion processing on the electronic medical record information based on the target feature data to obtain corresponding target electronic medical record information;
[0067] Calling a pre-built cost prediction model; wherein the cost prediction model is generated by training a preset gradient boosting decision forest model based on medical record sample data, and the gradient boosting decision forest model is a model of gradient boosting decision tree combined with random forest;
[0068] The target electronic medical record information is predicted based on the cost prediction model to generate a medical cost prediction value corresponding to the target electronic medical record information.
[0069] Compared with the prior art, the embodiments of the present application have the following beneficial effects:
[0070] The present application first obtains the user's initial electronic medical record information; then pre-processes the initial electronic medical record information to obtain corresponding electronic medical record information; then extracts the user's personal information from the electronic medical record information, and obtains the user's province information from the personal information; subsequently, filters out target feature data that matches the province information from preset province feature data; further performs information expansion processing on the electronic medical record information based on the target feature data to obtain corresponding target electronic medical record information; finally, calls a pre-constructed cost prediction model, and performs prediction processing on the target electronic medical record information based on the cost prediction model to generate a medical cost prediction value corresponding to the target electronic medical record information. The present application obtains electronic medical record information by preprocessing the initial electronic medical record information obtained, then obtains the user's province information from the electronic medical record information, and filters out target feature data matching the province information from preset province feature data, and then performs information expansion processing on the electronic medical record information based on the target feature data to obtain the corresponding target electronic medical record information, thereby providing an accurate, consistent, rich and complete input data set for the subsequent application of the cost prediction model, so that the subsequent use of the cost prediction model to perform predictive processing on the target electronic medical record information that has undergone information expansion processing can quickly and accurately generate medical expense prediction values corresponding to the target electronic medical record information, effectively improving the processing efficiency and accuracy of medical expense prediction, and ensuring the accuracy of the generated medical expense prediction values. BRIEF DESCRIPTION OF THE DRAWINGS
[0071] In order to more clearly illustrate the scheme in the present application, a brief introduction is given below to the drawings required for use in the description of the embodiments of the present application. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0072] Figure 1 is an exemplary system architecture diagram to which the present application may be applied;
[0073] Figure 2 A flowchart of an embodiment of a data prediction method based on artificial intelligence according to the present application;
[0074] Figure 3is a structural schematic diagram of an embodiment of an artificial intelligence-based data prediction device according to the present application;
[0075] Figure 4 It is a structural diagram of an embodiment of a computer device according to the present application. DETAILED DESCRIPTION
[0076] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of the present application; the terms used in the specification of the application herein are only for the purpose of describing specific embodiments and are not intended to limit the present application; the terms "including" and "having" and any variations thereof in the specification and claims of the present application and the above-mentioned drawings are intended to cover non-exclusive inclusions. The terms "first", "second", etc. in the specification and claims of the present application or the above-mentioned drawings are used to distinguish different objects, not to describe a specific order.
[0077] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.
[0078] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings.
[0079] like Figure 1 As shown, the system architecture 100 may include a terminal device 101, a network 102 and a server 103. The terminal device 101 may be a laptop 1011, a tablet computer 1012 or a mobile phone 1013. The network 102 is used to provide a medium for a communication link between the terminal device 101 and the server 103. The network 102 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.
[0080] The user can use the terminal device 101 to interact with the server 103 through the network 102 to receive or send messages, etc. Various communication client applications can be installed on the terminal device 101, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.
[0081] The terminal device 101 can be various electronic devices with a display screen and supporting web browsing. In addition to the laptop computer 1011, the tablet computer 1012 or the mobile phone 1013, the terminal device 101 can also be an e-book reader, an MP3 player (Moving Picture Experts Group Audio Layer III), an MP4 (Moving Picture Experts Group Audio Layer IV) player, a laptop computer, a desktop computer, etc.
[0082] The server 103 may be a server that provides various services, such as a background server that provides support for the page displayed on the terminal device 101. It should be noted that the artificial intelligence-based data prediction method provided in the embodiment of the present application is generally executed by a server / terminal device, and accordingly, the artificial intelligence-based data prediction device is generally set in the server / terminal device.
[0083] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.
[0084] Continue to refer Figure 2 , showing a flow chart of an embodiment of an artificial intelligence-based data prediction method according to the present application. According to different needs, the order of the steps in the flow chart can be changed, and some steps can be omitted. The artificial intelligence-based data prediction method provided in the embodiment of the present application can be applied to any scenario where medical expenses are required, and the artificial intelligence-based data prediction method can be applied to products in these scenarios, for example, medical expenses in the field of digital medicine. The artificial intelligence-based data prediction method comprises the following steps:
[0085] Step S201, obtaining the user's initial electronic medical record information.
[0086] In this embodiment, the data prediction method based on artificial intelligence is executed on the electronic device (e.g. Figure 1The server / terminal device shown in the figure) can obtain the user's initial electronic medical record information through a wired connection or a wireless connection. It should be noted that the above-mentioned wireless connection method may include but is not limited to 3G / 4G / 5G connection, Wi-Fi connection, Bluetooth connection, Wi MAX connection, Zigbee connection, UWB (Ultra Wi DeBand) connection, and other wireless connection methods currently known or to be developed in the future. The present application can be applied in the medical expense prediction scenario of medical insurance. In the medical expense prediction scenario, insurance products allow insurance with illness, which brings new challenges to insurance companies in risk assessment and pricing. For example, predict the different risk levels of insured persons with illnesses, predict the medical expenses such as medical expenses and hospitalization expenses that the insured needs to reimburse, and dynamically adjust the insurance premium rate according to the changes in the insured's health status. Among them, the initial electronic medical record information of the customer can be collected from the hospital channel, and the initial electronic medical record information may include the user's personal information, or personal basic information (such as age, gender, province, etc.), health status (such as medical history, physical examination results, etc.), and may also include other factors that may affect medical expenses.
[0087] Step S202: pre-process the initial electronic medical record information to obtain corresponding electronic medical record information.
[0088] In this embodiment, the above-mentioned specific implementation process of preprocessing the initial electronic medical record information to obtain the corresponding electronic medical record information will be further described in detail in subsequent specific embodiments of the present application, and will not be elaborated in detail here.
[0089] Step S203: extract the user's personal information from the electronic medical record information, and obtain the user's province information from the personal information.
[0090] In this embodiment, the user's personal information can be obtained by extracting personal information from the electronic medical record information, and then the province information can be extracted from the personal information to obtain the user's province information.
[0091] Step S204, filtering out target characteristic data matching the province information from the preset province characteristic data.
[0092] In this embodiment, the specific implementation process of filtering out the target characteristic data matching the province information from the preset province characteristic data will be described in further detail in subsequent specific embodiments of the present application and will not be elaborated on here.
[0093] Step S205: performing information expansion processing on the electronic medical record information based on the target feature data to obtain corresponding target electronic medical record information.
[0094] In this embodiment, the above-mentioned specific implementation process of performing information expansion processing on the electronic medical record information based on the target feature data to obtain the corresponding target electronic medical record information will be described in further details in subsequent specific embodiments of the present application and will not be elaborated in detail here.
[0095] Step S206, calling a pre-built cost prediction model.
[0096] In this embodiment, the cost prediction model is generated by training a preset gradient boosting decision forest model based on medical record sample data, and the gradient boosting decision forest model is a model of gradient boosting decision tree combined with random forest. The specific cost prediction model construction process will be further described in detail in the subsequent specific embodiments of this application, and will not be elaborated on here.
[0097] Step S207: performing prediction processing on the target electronic medical record information based on the cost prediction model to generate a medical cost prediction value corresponding to the target electronic medical record information.
[0098] In this embodiment, the target electronic medical record information can be input into the cost prediction model, and the target electronic medical record information can be predicted and processed by the cost prediction model, thereby outputting the prediction result corresponding to the target electronic medical record information, that is, the above-mentioned medical expense prediction value. Among them, the prediction result is usually output in numerical form. Furthermore, the predicted medical expenses can be used as an important reference factor for the pricing of insurance products. Based on the prediction results, insurance companies can evaluate the potential medical expense risks of different customer groups and formulate differentiated insurance premiums accordingly. And through actuarial analysis, it is ensured that the pricing of insurance products can not only cover the expected medical expenses, but also maintain the market competitiveness of the products.
[0099] The present application obtains electronic medical record information by preprocessing the initial electronic medical record information obtained, then obtains the user's province information from the electronic medical record information, and filters out target feature data matching the province information from preset province feature data, and then performs information expansion processing on the electronic medical record information based on the target feature data to obtain the corresponding target electronic medical record information, thereby providing an accurate, consistent, rich and complete input data set for the subsequent application of the cost prediction model, so that the subsequent use of the cost prediction model to perform predictive processing on the target electronic medical record information that has undergone information expansion processing can quickly and accurately generate medical expense prediction values corresponding to the target electronic medical record information, effectively improving the processing efficiency and accuracy of medical expense prediction, and ensuring the accuracy of the generated medical expense prediction values.
[0100] In some optional implementations, step S204 includes the following steps:
[0101] Get preset public statistics.
[0102] In this embodiment, the above-mentioned public statistical data may specifically include data sources such as the National Bureau of Statistics, insurance industry associations, and government reports.
[0103] The regional characteristic data and medical characteristic data of each province are extracted from the public statistical data to obtain the provincial characteristic data.
[0104] In this embodiment, the above-mentioned regional characteristic data may specifically include data such as insurance density, insurance depth, total number of insurance companies, social economic conditions (such as per capita GDP), etc. The above-mentioned medical characteristic data may specifically include data such as the medical care consumer price index, medical resource distribution (such as the number of medical and health institutions), etc.
[0105] Filter out designated regional characteristic data matching the provincial information from the regional characteristic data of each of the provinces.
[0106] In this embodiment, the regional characteristic data of each province is matched with the province information of the user to filter out the designated regional characteristic data matching the province information of the user.
[0107] Filter out designated medical characteristic data matching the provincial information from the medical characteristic data of each of the provinces.
[0108] In this embodiment, the medical characteristic data of each province is matched with the user's province information to filter out the designated medical characteristic data that matches the user's province information.
[0109] The target characteristic data is constructed based on the designated regional characteristic data and the designated medical characteristic data.
[0110] In this embodiment, the designated regional characteristic data and the designated medical characteristic data may be integrated to obtain corresponding integrated data, and the integrated data may be used as the target characteristic data.
[0111] This application obtains preset public statistical data; then extracts the regional characteristic data and medical characteristic data of each province from the public statistical data to obtain the provincial characteristic data; then filters out the designated regional characteristic data that matches the provincial information from the regional characteristic data of each province; and filters out the designated medical characteristic data that matches the provincial information from the medical characteristic data of each province; and subsequently constructs the target characteristic data based on the designated regional characteristic data and the designated medical characteristic data. This application extracts the regional characteristic data and medical characteristic data of each province from preset public statistical data, and then filters out the designated regional characteristic data and designated medical characteristic data that match the user's province information from the regional characteristic data and the medical characteristic data, and then based on the designated regional characteristic data and the designated medical characteristic data, it can realize the rapid and accurate construction of the target characteristic data, which is conducive to the subsequent accurate completion of the information expansion processing of the electronic medical record information according to the target characteristic data, and then the prediction accuracy of the subsequent use of the cost prediction model to predict the target electronic medical record information that has undergone information expansion processing.
[0112] In some optional implementations of this embodiment, step S205 includes the following steps:
[0113] Call the preset data analysis tool.
[0114] In this embodiment, the data analysis tool is a pre-built tool with functions such as merging and connecting.
[0115] The electronic medical record information and the target feature data are merged using a merge operation in the data analysis tool to obtain corresponding merged data.
[0116] In this embodiment, the electronic medical record information and the target feature data can be imported into the data analysis tool, and then a merge operation is used in the data analysis tool to merge the electronic medical record information and the target feature data, thereby obtaining corresponding merged data.
[0117] The combined data is used as the target feature data.
[0118] In this embodiment, the generated target characteristic data is data including the electronic medical record information and the target characteristic data.
[0119] This application calls a preset data analysis tool; then uses a merge operation in the data analysis tool to merge the electronic medical record information and the target feature data to obtain corresponding merged data; and subsequently uses the merged data as the target feature data. This application merges the electronic medical record information and the target feature data based on the use of a data analysis tool, so that the information expansion processing of the electronic medical record information can be completed quickly and accurately and the corresponding target feature data can be obtained, thereby improving the generation efficiency of the target feature data and ensuring the data accuracy of the obtained target feature data. This is conducive to the subsequent use of a cost prediction model to predict the target electronic medical record information that has undergone information expansion processing, which can effectively improve the accuracy of the prediction processing and improve the accuracy of the generated medical cost prediction value.
[0120] In some optional implementations, step S202 includes the following steps:
[0121] The initial electronic medical record information is cleaned to obtain corresponding first electronic medical record information.
[0122] In this embodiment, the above data cleaning includes: Missing value processing: filling missing data (such as using mean, median, mode, or filling through model prediction), or deleting records with too many missing values. Outlier processing: identifying and processing outliers in the data, which can be done by deleting, replacing, or adjusting according to business logic.
[0123] Get the preset classification encoding strategy.
[0124] In this embodiment, the strategy content of the above classification and coding strategy includes: Feature classification: classify the cleaned data (i.e., the first electronic medical record information) according to its meaning and nature, such as demographic characteristics, health characteristics, regional characteristics, etc. Feature encoding: encode the classification features so that the model can process them. One-hot encoding can be used for encoding processing. Alternatively, label encoding, target encoding, etc. can also be used.
[0125] The first electronic medical record information is encoded based on the classification coding strategy to obtain corresponding second electronic medical record information.
[0126] In this embodiment, the classification features in the first electronic medical record information can be encoded by using the classification encoding strategy, and the classification features can be converted into binary columns using OneHotEncoder, thereby obtaining the encoded second electronic medical record information. By encoding the first electronic medical record information, the amount of calculation of the model can be effectively reduced in the future.
[0127] The second electronic medical record information is used as the electronic medical record information.
[0128] The present application performs data cleaning processing on the initial electronic medical record information to obtain the corresponding first electronic medical record information; then obtains a preset classification and coding strategy; then encodes the first electronic medical record information based on the classification and coding strategy to obtain the corresponding second electronic medical record information; and subsequently uses the second electronic medical record information as the electronic medical record information. The present application performs data cleaning processing and encoding processing on the initial electronic medical record information, thereby realizing intelligent and accurate preprocessing of the initial electronic medical record information, which is beneficial for providing an accurate, consistent and complete input data set for the subsequent application of the cost prediction model, and can ensure the accuracy of the cost prediction model's prediction processing of the input data set.
[0129] In some optional implementations, before step S206, the electronic device may further perform the following steps:
[0130] Get pre-built sample medical records data.
[0131] In this embodiment, the electronic medical record information of the customer can be collected from hospital channels. The electronic medical record information may include basic personal information (such as age, gender, province, etc.), health status (such as medical history, physical examination results, etc.) and other factors that may affect medical expenses. Then, the collected electronic medical record information of the customer is preprocessed and information expanded, thereby completing the construction of the medical record sample data. Among them, the above-mentioned process of preprocessing and information expansion of the collected electronic medical record information of the customer can refer to the specific process of preprocessing the initial electronic medical record information of the user and expanding the information of the preprocessed electronic medical record information, which will not be described in detail here.
[0132] A random forest model is created, and the random forest model is trained using the medical record sample data to obtain a corresponding target random forest model.
[0133] In this embodiment, random forest is an integrated learning method that reduces the variance of the model and improves the accuracy of prediction by constructing multiple decision trees and performing voting or average prediction. The random forest model is trained using medical record sample data to obtain initial prediction results. The performance of the model can be optimized by adjusting the parameters of the random forest (such as the number of trees, maximum depth, etc.), thereby obtaining the corresponding target random forest model.
[0134] A gradient boosting decision forest model is created, and the target random forest model is used as an initial prediction estimator of the gradient boosting decision forest model.
[0135] In this embodiment, a model of a gradient boosting decision tree combined with a random forest is used as the above-mentioned gradient boosting decision forest model, and the target random forest model obtained by previous training is used as the initial prediction estimator of the gradient boosting decision forest model. Among them, the gradient boosting decision forest model gradually improves the model through multiple rounds of iterations, and the initial prediction estimator is a random forest model. Random forest is a bagging algorithm that uses a self-service sampling method to extract samples to form multiple sub-datasets, builds a decision tree for each sub-dataset, and randomly selects some features for splitting. In the decision tree of the figure, the circle represents the root node of the tree, and the square represents the leaf node of the tree. Random forest constructs multiple decision trees by introducing randomness, and obtains the final result by voting or averaging. Then train a new decision tree to reduce the residual of the previous round of model. The residual is the difference between the actual observation value and the current model prediction value. The new tree will learn how to correct these residuals and finally obtain the integrated prediction result. This process enables the model to capture the complex relationship in the data, thereby improving the prediction accuracy. In addition, the gradient boosting decision forest model samples and features to increase the independence between the initialization estimator random forest model. The gradient boosting decision forest model combines the Bagging and Boosting algorithms, significantly reducing the mean absolute error (MAE) and mean absolute percentage error (MAPE) of the model. It performs well in dealing with outliers and has a certain ability to resist overfitting.
[0136] Determine the target loss function and target hyperparameter tuning method corresponding to the gradient boosting decision forest model.
[0137] In this embodiment, the above-mentioned target loss function can specifically adopt the Huber loss function. The Huber loss function is a loss function that combines the mean square error and the absolute loss. It is a compromise between the mean square error and the absolute loss. For abnormal points far from the center, the absolute loss is used, while the points near the center use the average square error. The Huber loss function has a certain robustness to outliers. In the gradient boosting decision forest model, the prediction performance of the model is optimized by using the Huber loss function as the loss function of the model.
[0138] In addition, the above-mentioned target hyperparameter tuning method can specifically adopt the Bayesian optimization method. The optimal hyperparameter combination is found by using hyperparameter tuning methods such as Bayesian optimization. Bayesian optimization is an optimization method based on Bayes' theorem, which can find the global optimal solution or approximate optimal solution under limited computing resources. Among them, by adjusting the hyperparameters of the gradient boosting decision forest model (such as learning rate, number of trees, maximum depth, etc.), the performance of the model can be further optimized.
[0139] Based on the target loss function and the target hyperparameter tuning method, the gradient boosting decision forest model is trained with ten-fold cross validation using the medical record sample data to obtain a specified gradient boosting decision forest model that meets the preset model building conditions.
[0140] In this embodiment, the process of performing ten-fold cross validation training on the gradient boosting decision forest model specifically includes: ten-fold cross validation division: randomly shuffle the medical record sample data to ensure the consistency of data distribution. Then divide the data set into ten parts (if the data set size is not a multiple of 10, some parts may contain slightly more or less data points). Each part of the data is used as a validation set in turn, and the remaining nine parts are used as training sets. Ten-fold cross validation training: for each fold cross validation: use all data other than the currently specified validation set as the training set. Use the training set data to train the gradient boosting decision forest model. Evaluate the performance of the model on the current validation set, but do not update the model parameters. Repeat the above steps until each part of the data is used as a validation set. Performance evaluation: 1. Based on calculating the average training l oss: In this step, the "average training l oss" is usually not directly calculated, because in the ten-fold cross validation, the training set for each round is different. However, the l oss in each round of training can be recorded for analysis. If necessary, the average value of l oss in all training rounds can be calculated, but this does not directly reflect the performance of the model on unseen data. 2. Calculate the average validation loss: For each round of cross-validation, record the loss of the model on the validation set. Calculate the average of these loss, which reflects the average performance of the model on unseen data. 3. Evaluate model performance: In addition to loss, other performance indicators can be calculated, such as accuracy (for classification problems), R2 score (for regression problems), etc. Analyze the average validation loss and performance indicators to evaluate the stability and generalization ability of the model. 4. Adjustment and optimization: Based on the results of the performance evaluation, use hyperparameter tuning methods such as Bayesian optimization to adjust the parameters of the gradient boosting decision forest model (such as learning rate, number of trees, maximum depth, etc.). Repeat model training and validation until the optimal combination of model parameters is found. 5. Final model selection: Select the model that performs best on the validation set as the final model. This final model can be retrained using the entire dataset (including the data previously used as the validation set) to obtain better performance (but this may introduce some risk of overfitting, so it needs to be handled with caution).
[0141] The specified gradient boosting decision forest model is used as the cost prediction model.
[0142] This application uses pre-constructed medical record sample data to train the random forest model to obtain a corresponding target random forest model, and uses the target random forest model as the initial prediction estimator of the pre-created gradient boosting decision forest model. Then, based on the determined target loss function and target hyperparameter tuning method corresponding to the gradient boosting decision forest model, the medical record sample data is used to perform ten-fold cross-validation training on the gradient boosting decision forest model to obtain a specified gradient boosting decision forest model that meets the preset model construction conditions and serves as the final cost prediction model, thereby effectively improving the construction efficiency of the cost prediction model and ensuring the model effect of the obtained cost prediction model.
[0143] In some optional implementations of this embodiment, after the step of using the specified gradient boosting decision forest model as the cost prediction model, the electronic device may further perform the following steps:
[0144] Obtain a specified storage strategy corresponding to the cost prediction model.
[0145] In this embodiment, the above-mentioned specified storage strategy may specifically adopt strategies such as cyclic overwriting or scheduled backup to ensure that the cost prediction model will not be lost due to insufficient storage space.
[0146] A designated storage medium corresponding to the cost prediction model is determined.
[0147] In this embodiment, there is no specific limitation on the selection of the above-mentioned designated storage medium, and it may include, for example, a local database, a hard disk, a cloud server, a blockchain, etc.
[0148] The cost prediction model is stored based on the specified storage strategy and the specified storage medium.
[0149] In this embodiment, the cost prediction model can be stored in the designated storage medium in a storage manner with a designated storage strategy.
[0150] The present application obtains the designated storage strategy corresponding to the cost prediction model; then determines the designated storage medium corresponding to the cost prediction model; and subsequently performs storage processing on the cost prediction model based on the designated storage strategy and the designated storage medium. After completing the construction of the cost prediction model, the present application intelligently obtains the designated storage strategy corresponding to the cost prediction model, determines the designated storage medium corresponding to the cost prediction model, and then performs storage processing on the cost prediction model based on the designated storage strategy and the designated storage medium, thereby ensuring the data security of the cost prediction model and ensuring that the cost prediction model will not be lost due to insufficient storage space, thereby improving the storage intelligence of the cost prediction model.
[0151] In some optional implementations of this embodiment, after step S207, the electronic device may further perform the following steps:
[0152] The communication information of the user is obtained based on the personal information.
[0153] In this embodiment, the communication information of the user can be obtained by extracting the communication information from the personal information. The communication information may include the user's mobile phone number, email address, etc.
[0154] Obtain explanation information corresponding to the predicted value of medical expenses.
[0155] In this embodiment, the above explanation information is pre-generated information for explaining how the medical expense prediction value is generated. Exemplarily, the content of the explanation information may include: the medical expense prediction value generated this time is predicted and generated based on the expense prediction model for your electronic medical record information, please use it for reference, thank you.
[0156] A corresponding medical expense prediction report is generated based on the medical expense prediction value and the explanation information.
[0157] In this embodiment, the medical expense forecast value and the explanation information can be filled into the corresponding positions in the preset medical expense forecast report template to obtain the corresponding medical expense forecast report. The template content of the above medical expense forecast report template can be set according to the actual report construction requirements, for example, it can at least include a cost area, an explanation area, etc.
[0158] Based on the communication information, the medical expense prediction report is sent to the user.
[0159] In this embodiment, the medical expense prediction report can be sent to the communication terminal of the corresponding user through the above communication information.
[0160] This application obtains the user's communication information based on the personal information; then obtains the explanatory information corresponding to the medical expense prediction value; then generates a corresponding medical expense prediction report based on the medical expense prediction value and the explanatory information; and subsequently sends the medical expense prediction report to the user based on the communication information. After the application predicts and processes the target electronic medical record information based on the expense prediction model and generates a medical expense prediction value corresponding to the target electronic medical record information, it will also intelligently obtain the explanatory information corresponding to the medical expense prediction value, and generate a corresponding medical expense prediction report based on the medical expense prediction value and the explanatory information, and then send the medical expense prediction report to the user based on the user's communication information obtained from the personal information, so that the user can easily understand the medical expense prediction value corresponding to his or her own electronic medical record information, and based on the use of the explanatory information, the user can know the origin of the generation of the medical expense prediction value, which helps to improve the user's experience.
[0161] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.
[0162] It should be emphasized that in order to further ensure the privacy and security of the above-mentioned medical expense prediction values, the above-mentioned medical expense prediction values can also be stored in a node of a blockchain.
[0163] The blockchain referred to in this application is a new application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, encryption algorithm, etc. Blockchain is essentially a decentralized database, a string of data blocks generated by cryptographic methods. Each data block contains a batch of network transaction information, which is used to verify the validity of its information (anti-counterfeiting) and generate the next block. Blockchain can include the underlying blockchain platform, platform product service layer, and application service layer.
[0164] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Artificial Intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.
[0165] AI basic technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. AI software technologies mainly include computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning.
[0166] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through computer-readable instructions, and the computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the aforementioned storage medium can be a non-volatile storage medium such as a disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).
[0167] It should be understood that, although the steps in the flowchart of the accompanying drawings are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.
[0168] Further references Figure 3 , as a response to the above Figure 2 The present application provides an embodiment of a data prediction device based on artificial intelligence. Figure 2 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.
[0169] like Figure 3 As shown, the artificial intelligence-based data prediction device 300 described in this embodiment includes: a first acquisition module 301, a preprocessing module 302, an extraction module 303, a screening module 304, an expansion module 305, a calling module 306 and a prediction module 307. Among them:
[0170] The first acquisition module 301 is used to acquire the user's initial electronic medical record information;
[0171] A preprocessing module 302 is used to preprocess the initial electronic medical record information to obtain corresponding electronic medical record information;
[0172] An extraction module 303 is used to extract the user's personal information from the electronic medical record information, and obtain the user's province information from the personal information;
[0173] A screening module 304 is used to screen out target characteristic data matching the province information from the preset province characteristic data;
[0174] An expansion module 305 is used to perform information expansion processing on the electronic medical record information based on the target feature data to obtain corresponding target electronic medical record information;
[0175] A calling module 306 is used to call a pre-built cost prediction model; wherein the cost prediction model is generated by training a preset gradient boosting decision forest model based on medical record sample data, and the gradient boosting decision forest model is a model of gradient boosting decision tree combined with random forest;
[0176] The prediction module 307 is used to perform prediction processing on the target electronic medical record information based on the cost prediction model to generate a medical cost prediction value corresponding to the target electronic medical record information.
[0177] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the artificial intelligence-based data prediction method in the aforementioned implementation mode, and will not be repeated here.
[0178] In some optional implementations of this embodiment, the screening module 304 includes:
[0179] The first acquisition submodule is used to acquire preset public statistical data;
[0180] An extraction submodule, used to extract the regional characteristic data and medical characteristic data of each province from the public statistical data to obtain the provincial characteristic data;
[0181] The first screening submodule is used to screen out the designated regional characteristic data matching the provincial information from the regional characteristic data of each province;
[0182] The second screening submodule is used to screen out the designated medical characteristic data matching the provincial information from the medical characteristic data of each province;
[0183] A construction submodule is used to construct the target characteristic data based on the designated regional characteristic data and the designated medical characteristic data.
[0184] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the artificial intelligence-based data prediction method in the aforementioned implementation mode, and will not be repeated here.
[0185] In some optional implementations of this embodiment, the expansion module 305 includes:
[0186] Call submodule, used to call preset data analysis tools;
[0187] A merging submodule, used to merge the electronic medical record information and the target feature data using a merging operation in the data analysis tool to obtain corresponding merged data;
[0188] The first determining submodule is used to use the combined data as the target feature data.
[0189] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the artificial intelligence-based data prediction method in the aforementioned implementation mode, and will not be repeated here.
[0190] In some optional implementations of this embodiment, the preprocessing module 302 includes:
[0191] A cleaning submodule, used for performing data cleaning processing on the initial electronic medical record information to obtain corresponding first electronic medical record information;
[0192] The second acquisition submodule is used to acquire a preset classification coding strategy;
[0193] An encoding submodule, used for encoding the first electronic medical record information based on the classification encoding strategy to obtain corresponding second electronic medical record information;
[0194] The second determining submodule is used to use the second electronic medical record information as the electronic medical record information.
[0195] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the artificial intelligence-based data prediction method in the aforementioned implementation mode, and will not be repeated here.
[0196] In some optional implementations of this embodiment, the data prediction device based on artificial intelligence further includes:
[0197] A second acquisition module is used to acquire pre-constructed medical record sample data;
[0198] A first creation module is used to create a random forest model, and use the medical record sample data to train the random forest model to obtain a corresponding target random forest model;
[0199] A second creation module is used to create a gradient boosting decision forest model and use the target random forest model as an initial prediction estimator of the gradient boosting decision forest model;
[0200] A first determination module is used to determine a target loss function and a target hyperparameter tuning method corresponding to the gradient boosting decision forest model;
[0201] A training module, for performing a ten-fold cross validation training on the gradient boosting decision forest model using the medical record sample data based on the target loss function and the target hyperparameter tuning method, to obtain a specified gradient boosting decision forest model that meets the preset model building conditions;
[0202] The second determination module is used to use the specified gradient boosting decision forest model as the cost prediction model.
[0203] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the artificial intelligence-based data prediction method in the aforementioned implementation mode, and will not be repeated here.
[0204] In some optional implementations of this embodiment, the data prediction device based on artificial intelligence further includes:
[0205] A third acquisition module is used to acquire a specified storage strategy corresponding to the cost prediction model;
[0206] A third determination module, used to determine a designated storage medium corresponding to the cost prediction model;
[0207] A storage module is used to store the cost prediction model based on the specified storage strategy and the specified storage medium.
[0208] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the artificial intelligence-based data prediction method in the aforementioned implementation mode, and will not be repeated here.
[0209] In some optional implementations of this embodiment, the data prediction device based on artificial intelligence further includes:
[0210] A fourth acquisition module, used to acquire the communication information of the user based on the personal information;
[0211] A fifth acquisition module, used to acquire explanation information corresponding to the predicted value of medical expenses;
[0212] A generating module, used for generating a corresponding medical expense prediction report based on the medical expense prediction value and the explanation information;
[0213] A sending module is used to send the medical expense prediction report to the user based on the communication information.
[0214] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the artificial intelligence-based data prediction method in the aforementioned implementation mode, and will not be repeated here.
[0215] To solve the above technical problems, the present application also provides a computer device. Figure 4 , Figure 4 This is a basic structural block diagram of the computer device in this embodiment.
[0216] The computer device 4 includes a memory 41, a processor 42, and a network interface 43 that are interconnected through a system bus. It should be noted that the figure only shows a computer device 4 with components 41-43, but it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented instead. Among them, those skilled in the art can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital processors (DSPs), embedded devices, etc.
[0217] The computer device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The computer device may interact with a user through a keyboard, a mouse, a remote controller, a touch pad, or a voice control device.
[0218] The memory 41 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, etc. In some embodiments, the memory 41 can be an internal storage unit of the computer device 4, such as a hard disk or memory of the computer device 4. In other embodiments, the memory 41 can also be an external storage device of the computer device 4, such as a plug-in hard disk, a smart memory card (Smart Med ia Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the computer device 4. Of course, the memory 41 can also include both the internal storage unit of the computer device 4 and its external storage device. In this embodiment, the memory 41 is generally used to store the operating system and various application software installed on the computer device 4, such as computer-readable instructions of a data prediction method based on artificial intelligence, etc. In addition, the memory 41 can also be used to temporarily store various types of data that have been output or are to be output.
[0219] The processor 42 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips in some embodiments. The processor 42 is generally used to control the overall operation of the computer device 4. In this embodiment, the processor 42 is used to run the computer-readable instructions stored in the memory 41 or process data, such as running the computer-readable instructions of the artificial intelligence-based data prediction method.
[0220] The network interface 43 may include a wireless network interface or a wired network interface. The network interface 43 is generally used to establish a communication connection between the computer device 4 and other electronic devices.
[0221] The present application also provides another embodiment, namely, providing a computer-readable storage medium, wherein the computer-readable storage medium stores computer-readable instructions, and the computer-readable instructions can be executed by at least one processor to enable the at least one processor to perform the steps of the artificial intelligence-based data prediction method as described above.
[0222] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present application.
[0223] Obviously, the embodiments described above are only some embodiments of the present application, rather than all embodiments. The preferred embodiments of the present application are given in the accompanying drawings, but they do not limit the patent scope of the present application. The present application can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present application more thorough and comprehensive. Although the present application is described in detail with reference to the aforementioned embodiments, for those skilled in the art, it is still possible to modify the technical solutions recorded in the aforementioned specific implementation methods, or to perform equivalent replacement of some of the technical features therein. Any equivalent structure made using the contents of the specification and drawings of this application, directly or indirectly used in other related technical fields, is similarly within the scope of patent protection of this application.
Claims
1. A data prediction method based on artificial intelligence, characterized in that: The steps include: Obtain the user's initial electronic medical record information; Preprocessing the initial electronic medical record information to obtain corresponding electronic medical record information; Extracting the user's personal information from the electronic medical record information, and obtaining the user's province information from the personal information; Filtering target characteristic data matching the province information from the preset province characteristic data; Performing information expansion processing on the electronic medical record information based on the target feature data to obtain corresponding target electronic medical record information; Calling a pre-built cost prediction model; wherein the cost prediction model is generated by training a preset gradient boosting decision forest model based on medical record sample data, and the gradient boosting decision forest model is a model of gradient boosting decision tree combined with random forest; The target electronic medical record information is predicted based on the cost prediction model to generate a medical cost prediction value corresponding to the target electronic medical record information.
2. The artificial intelligence-based data prediction method according to claim 1, characterized in that: The step of selecting target characteristic data matching the province information from the preset province characteristic data specifically includes: Get preset public statistics; Extracting the regional characteristic data and medical characteristic data of each province from the public statistical data to obtain the provincial characteristic data; Filter out the designated regional characteristic data matching the provincial information from the regional characteristic data of each province; Filter out designated medical characteristic data matching the provincial information from the medical characteristic data of each of the provinces; The target characteristic data is constructed based on the designated regional characteristic data and the designated medical characteristic data.
3. The data prediction method based on artificial intelligence according to claim 1, characterized in that: The step of performing information expansion processing on the electronic medical record information based on the target feature data to obtain corresponding target electronic medical record information specifically includes: Call the preset data analysis tool; Using a merge operation in the data analysis tool to merge the electronic medical record information with the target feature data to obtain corresponding merged data; The combined data is used as the target feature data.
4. The data prediction method based on artificial intelligence according to claim 1, characterized in that: The step of preprocessing the initial electronic medical record information to obtain corresponding electronic medical record information specifically includes: Performing data cleaning on the initial electronic medical record information to obtain corresponding first electronic medical record information; Obtain the preset classification coding strategy; Encoding the first electronic medical record information based on the classification coding strategy to obtain corresponding second electronic medical record information; The second electronic medical record information is used as the electronic medical record information.
5. The artificial intelligence-based data prediction method according to claim 1, characterized in that: Before the step of calling the pre-built cost prediction model, the method further includes: Get pre-built sample medical record data; Creating a random forest model, and using the medical record sample data to train the random forest model to obtain a corresponding target random forest model; Creating a gradient boosting decision forest model and using the target random forest model as an initial prediction estimator for the gradient boosting decision forest model; Determine a target loss function and a target hyperparameter tuning method corresponding to the gradient boosting decision forest model; Based on the target loss function and the target hyperparameter tuning method, the gradient boosting decision forest model is trained with ten-fold cross validation using the medical record sample data to obtain a specified gradient boosting decision forest model that meets the preset model building conditions; The specified gradient boosting decision forest model is used as the cost prediction model.
6. The artificial intelligence-based data prediction method according to claim 5, characterized in that: After the step of specifying the gradient boosting decision forest model as the cost prediction model, the method further includes: Obtaining a specified storage strategy corresponding to the cost prediction model; Determining a designated storage medium corresponding to the cost prediction model; The cost prediction model is stored based on the specified storage strategy and the specified storage medium.
7. The artificial intelligence-based data prediction method according to claim 1, characterized in that: After the step of performing prediction processing on the target electronic medical record information based on the cost prediction model to generate a medical cost prediction value corresponding to the target electronic medical record information, the method further includes: Obtaining communication information of the user based on the personal information; Obtaining explanation information corresponding to the predicted value of medical expenses; Generate a corresponding medical expense prediction report based on the medical expense prediction value and the explanation information; Based on the communication information, the medical expense prediction report is sent to the user.
8. A data prediction device based on artificial intelligence, characterized in that: include: The first acquisition module is used to acquire the user's initial electronic medical record information; A preprocessing module, used to preprocess the initial electronic medical record information to obtain corresponding electronic medical record information; An extraction module, used to extract the user's personal information from the electronic medical record information, and obtain the user's province information from the personal information; A screening module, used to screen out target characteristic data matching the province information from preset province characteristic data; An expansion module, used to perform information expansion processing on the electronic medical record information based on the target feature data to obtain corresponding target electronic medical record information; A calling module, used to call a pre-built cost prediction model; wherein the cost prediction model is generated by training a preset gradient boosting decision forest model based on medical record sample data, and the gradient boosting decision forest model is a model of a gradient boosting decision tree combined with a random forest; The prediction module is used to perform prediction processing on the target electronic medical record information based on the cost prediction model to generate a medical cost prediction value corresponding to the target electronic medical record information.
9. A computer device, characterized in that: It comprises a memory and a processor, wherein the memory stores computer-readable instructions, and when the processor executes the computer-readable instructions, the steps of the artificial intelligence-based data prediction method as described in any one of claims 1 to 7 are implemented.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the artificial intelligence-based data prediction method according to any one of claims 1 to 7.
