Resource allocation method, apparatus, device, and storage medium

Abstract

Embodiments of the application disclose a resource allocation method, device, equipment and storage medium, and belong to the technical field of artificial intelligence and data processing. The method comprises: acquiring input parameters related to medical institution visit frequency prediction; determining k groups of random parameters based on the input parameters and target correlation, k being a positive integer; performing k rounds of visit frequency prediction processing based on the input parameters and the k groups of random parameters to obtain k groups of initial visit frequency prediction results of the medical institution; determining a final visit frequency prediction result of the medical institution based on the k groups of initial visit frequency prediction results; and determining resource allocation suggestion information for the medical institution based on the final visit frequency prediction result of the medical institution. The embodiments of the application make the resource allocation of the medical institution more reasonable and achieve maximum utilization efficiency of resources.

Description

Technical Field

[0001] This application relates to the fields of artificial intelligence and data processing technology, and in particular to a resource allocation method, apparatus, device and storage medium. Background Technology

[0002] Medical insurance is a social insurance system established to compensate workers for economic losses caused by the risk of illness.

[0003] In this technology, users can seek medical treatment at different types of medical institutions, each with different medical insurance reimbursement policies. Some users, even with a simple cold, go to primary-level medical institutions, leading to an overabundance of patients at these institutions and an underabundance at tertiary-level medical institutions. This results in an unreasonable allocation of medical resources and an inability to maximize resource utilization efficiency. Summary of the Invention

[0004] This application provides a resource allocation method, apparatus, device, and storage medium, enabling more rational resource allocation in medical institutions and maximizing resource utilization efficiency. The technical solution is as follows:

[0005] On one hand, embodiments of this application provide a resource allocation method, the method comprising:

[0006] Obtain input parameters related to the prediction of the number of outpatient visits to the medical institution, wherein the input parameters refer to user-input parameters related to the prediction of the number of outpatient visits to the medical institution;

[0007] Based on the input parameters and target correlation, k sets of random parameters are determined. The random parameters refer to randomly generated parameters that are related to the predicted number of outpatient visits to the medical institution. The target correlation includes the correlation between the random parameters and the input parameters. The k is a positive integer.

[0008] Based on the input parameters and the k sets of random parameters, k rounds of outpatient visit prediction processing are performed to obtain the initial outpatient visit prediction results for the medical institution.

[0009] Based on the initial patient visit prediction results of the k groups, the final patient visit prediction results of the medical institution are determined.

[0010] Based on the final predicted number of outpatient visits to the medical institution, resource allocation recommendations are determined for the medical institution. These recommendations are used to provide suggestions when allocating resources to the medical institution.

[0011] On the other hand, embodiments of this application provide a resource allocation device, the device comprising:

[0012] The parameter acquisition module is used to acquire input parameters related to the prediction of the number of outpatient visits to the medical institution. The input parameters refer to the user-input parameters related to the prediction of the number of outpatient visits to the medical institution.

[0013] The parameter generation module is used to determine k sets of random parameters based on the input parameters and the target correlation. The random parameters refer to randomly generated parameters related to the predicted number of outpatient visits of the medical institution. The target correlation includes the correlation between the random parameters and the input parameters. The k is a positive integer.

[0014] The parameter processing module is used to perform k rounds of visitor prediction processing based on the input parameters and the k sets of random parameters to obtain the k sets of initial visitor prediction results for the medical institution.

[0015] The parameter processing module is also used to determine the final predicted number of visits to the medical institution based on the initial predicted number of visits for the k groups.

[0016] The information determination module is used to determine resource allocation suggestions for the medical institution based on the final predicted number of outpatient visits. The resource allocation suggestions are used to provide recommendations when allocating resources to the medical institution.

[0017] On the other hand, embodiments of this application provide a computer device, the computer device including a processor and a memory, the memory storing at least one instruction, at least one program, code set or instruction set, the at least one instruction, the at least one program, the code set or instruction set being loaded and executed by the processor to implement the resource allocation method as described above.

[0018] In another aspect, embodiments of this application provide a computer-readable storage medium storing at least one instruction, at least one program, a code set, or an instruction set, wherein the at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by a processor to implement the resource allocation method as described above.

[0019] In another aspect, embodiments of this application provide a computer program product or computer program that includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the resource allocation method provided in the above aspects.

[0020] The technical solution provided in this application can bring the following beneficial effects:

[0021] By determining the values ​​of multiple sets of random parameters based on the correlation between input parameters and random parameters, and then determining multiple initial predicted visitor counts based on the input parameters and multiple sets of random parameters, the final predicted visitor count is determined based on the multiple initial predicted visitor counts. This embodiment of the application considers the correlation between various parameters and assigns values ​​to random parameters based on this correlation. It obtains the distribution of possible result ranges (i.e., the above multiple sets of initial predicted visitor counts) through random simulation, rather than just estimating the result at a single point. Therefore, it considers the volatility of medium- and long-term predictions and improves the accuracy of the final predicted visitor count. Based on the final predicted visitor count for the medical institution, resource allocation recommendations are determined for the medical institution. Since visitor counts are fully considered in resource allocation, the resource allocation of the medical institution is more reasonable, and the efficiency of resource utilization is maximized. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a schematic diagram of an implementation environment provided in one embodiment of this application;

[0024] Figure 2 This is a flowchart of a resource allocation method provided in one embodiment of this application;

[0025] Figure 3 This is a flowchart of a resource allocation method provided in another embodiment of this application;

[0026] Figure 4 This is a schematic diagram of the framework of a DFA model provided in one embodiment of this application;

[0027] Figure 5 This is a schematic diagram illustrating the value at risk provided in one embodiment of this application;

[0028] Figure 6 This is a flowchart of a resource allocation method provided in another embodiment of this application;

[0029] Figure 7 This is a schematic diagram of a prediction interface provided in one embodiment of this application;

[0030] Figure 8 This is a block diagram of a resource allocation apparatus provided in one embodiment of this application;

[0031] Figure 9This is a schematic diagram of the structure of a computer device provided in one embodiment of this application. Detailed Implementation

[0032] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0033] First, let me introduce and explain the relevant terms used in this application:

[0034] Medical Insurance Fund: Abbreviated as Medical Insurance Fund. The Medical Insurance Fund refers to a special fund raised by the state from employers and individuals in accordance with relevant national regulations to guarantee basic medical care for employees. It consists of two parts: a social pooling fund and individual accounts, jointly contributed by employers and employees at a certain ratio. The Medical Insurance Fund is one of the five major social insurance programs in my country. It is a monetary fund collected through laws or contracts from participating enterprises, institutions, organizations, or individuals at a predetermined ratio, providing basic medical security for the insured. All basic medical insurance premiums paid by insured individuals are deposited into their individual accounts; the premiums paid by participating employers, except for the portion deposited into individual accounts, are all allocated to the pooling fund, managed and allocated uniformly by the medical insurance agency. Individual accounts are mainly used to pay for outpatient and medication expenses, while the pooling fund is mainly used to pay for inpatient, outpatient specific items, and medical expenses for certain chronic diseases and home-based care. The Major Illness Medical Assistance Fund is used to pay for inpatient medical expenses exceeding the maximum payment limit of the basic medical insurance pooling fund.

[0035] Dynamic Financial Analysis (DFA) is a holistic financial modeling approach that simulates a company's future operating environment and results to demonstrate how changes in the external environment and internal strategic decisions affect a company's performance. It enables comprehensive forecasting of a company's operations and financial condition, and allows for dynamic monitoring of changes in its assets and liabilities, thus building an effective financial risk early warning system.

[0036] VAR (Value at Risk): The maximum possible loss of a financial asset or portfolio of securities under normal market fluctuations.

[0037] Artificial intelligence (AI) is the theory, methods, technology, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to achieve optimal results. In other words, AI is a comprehensive technology within computer science that attempts to understand the essence of intelligence and produce a new kind of intelligent machine that can react in a way similar to human intelligence. AI studies the design principles and implementation methods of various intelligent machines, enabling them to possess the functions of perception, reasoning, and decision-making.

[0038] Artificial intelligence (AI) is a comprehensive discipline encompassing a wide range of fields, including both hardware and software technologies. Fundamental AI technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies primarily include computer vision, speech processing, natural language processing, and machine learning / deep learning.

[0039] Machine Learning (ML) is a multidisciplinary field involving probability theory, statistics, approximation theory, convex analysis, and algorithm complexity theory. It specifically studies how computers can simulate or implement human learning behavior to acquire new knowledge or skills and reorganize existing knowledge structures to continuously improve their performance. Machine learning is the core of artificial intelligence and the fundamental way to endow computers with intelligence; its applications span all areas of artificial intelligence. Machine learning and deep learning typically include techniques such as artificial neural networks, belief networks, reinforcement learning, transfer learning, inductive learning, and instruction-based learning.

[0040] With the research and advancement of artificial intelligence (AI) technology, AI is being studied and applied in various fields, such as smart homes, smart wearable devices, virtual assistants, smart speakers, smart marketing, autonomous driving, drones, robots, smart healthcare, and smart customer service. It is believed that with the development of technology, AI will be applied in more fields and play an increasingly important role.

[0041] The solutions provided in this application involve technologies such as machine learning in artificial intelligence, and are specifically illustrated through the following embodiments.

[0042] Please refer to Figure 1 The diagram illustrates an implementation environment provided in one embodiment of this application. This implementation environment may include: terminal 10 and server 20.

[0043] In this embodiment, terminal 10 refers to a device that acquires parameters input by the user related to the prediction of the number of visits to a medical institution. Exemplarily, terminal 10 can be an electronic device such as a mobile phone, tablet computer, PC (Personal Computer), or smart wearable device, and this embodiment does not limit it to this.

[0044] In this embodiment, server 20 refers to the device that determines the final predicted number of outpatient visits to the medical institution. Exemplarily, server 20 can be a single server, a server cluster consisting of multiple servers, or a cloud server; this embodiment does not limit the specific type of server.

[0045] For example, terminal 10 and server 20 can communicate with each other, for example, through a wired network or a wireless network.

[0046] In one possible implementation, after obtaining the input parameters related to the predicted number of outpatient visits to the medical institution, terminal 10 sends them to server 20. Server 20, upon receiving the input parameters, determines k sets of random parameters based on target relevance. Then, it performs k rounds of outpatient visit prediction processing based on the input parameters and the k sets of random parameters to obtain k initial outpatient visit prediction results for the medical institution. Based on these k initial outpatient visit prediction results, it determines the final outpatient visit prediction result for the medical institution. Based on the final outpatient visit prediction result for the medical institution, it determines resource allocation recommendations for the medical institution. In another possible implementation, after determining the resource allocation recommendations for the medical institution, the server sends them to terminal 10, allowing terminal 10 to display the resource allocation recommendations on its interface. In yet another possible implementation, after server 20 determines the final outpatient visit prediction result for the medical institution, it sends the final outpatient visit prediction result to terminal 10, allowing terminal 10 to determine resource allocation recommendations for the medical institution based on the final outpatient visit prediction result.

[0047] In one possible implementation, the final predicted number of outpatient visits for the medical institution can be determined by terminal 10. In this case, terminal 10 acquires input parameters related to the predicted number of outpatient visits for the medical institution; determines k sets of random parameters based on target relevance; performs k rounds of outpatient visit prediction processing based on the input parameters and the k sets of random parameters to obtain k sets of initial outpatient visit prediction results for the medical institution; and determines the final predicted number of outpatient visits for the medical institution based on the k sets of initial outpatient visit prediction results. For example, terminal 10 determines resource allocation recommendations for the medical institution based on the final outpatient visit prediction results and displays these recommendations on the interface.

[0048] For ease of explanation, this application will use the example of a computer device determining resource allocation recommendations for medical institutions to illustrate this application. The computer device refers to an electronic device with computing and processing capabilities, including terminals or servers.

[0049] Please refer to Figure 2 The diagram illustrates a flowchart of a resource allocation method provided in one embodiment of this application. This method can be executed by a computer device and may include the following steps.

[0050] Step 201: Obtain input parameters related to the prediction of the number of visits to medical institutions.

[0051] In this embodiment of the application, the input parameters refer to user-input parameters related to the prediction of the number of visits to medical institutions.

[0052] In one example, the user enters input parameters related to the predicted number of visits to the medical institution in the prediction interface of the terminal. After the terminal obtains the input parameters, it sends them to the server so that the server can perform the following data processing flow.

[0053] In another example, the user enters input parameters related to the predicted number of visits to medical institutions in the prediction interface of the terminal. After the terminal obtains the input parameters, it executes the following data processing flow.

[0054] In possible implementations, the type of input parameters can be default or determined by the user according to requirements. Input parameters may include multiple parameters or only one parameter; this application embodiment does not limit the number of parameters included in the input parameters.

[0055] In one possible implementation, the computer device acquires the values ​​of input parameters related to the prediction of patient visits at the medical institution.

[0056] In possible implementations, the medical institution can be of any level. For example, the medical institution in the embodiments of this application can be a first type of medical institution (level 1 medical institution), a second type of medical institution (level 2 medical institution), a third type of medical institution (level 3 medical institution), etc.

[0057] Step 202: Based on the input parameters and the target correlation, determine k sets of random parameters, where k is a positive integer.

[0058] In this embodiment, the random parameter refers to a randomly generated parameter related to the prediction of patient visits to a medical institution. For example, the random parameter may include the number of annual patient visits by an individual within the i-th age range who visits a medical institution of the j-th type. Exemplarily, the number of annual patient visits by an individual follows a Poisson distribution.

[0059] In this embodiment, the target correlation includes the correlation between random parameters and input parameters. In possible implementations, the target correlation may also include the correlation between random parameters and the correlation between random parameters.

[0060] In one possible implementation, the computer device determines the values ​​of k sets of random parameters based on the values ​​of the input parameters and the relevance to the target.

[0061] Step 203: Based on the input parameters and k sets of random parameters, perform k rounds of patient visit prediction processing to obtain the initial patient visit prediction results for the medical institution.

[0062] The initial patient visit prediction result is used to indicate the patient visit prediction result obtained in each simulation. The patient visit prediction result refers to the predicted number of patients to be treated at this medical institution in the future.

[0063] Each set of random parameters corresponds to one simulation. Based on the input parameters and k sets of random parameters, k rounds of patient visit prediction processing are performed to obtain k initial patient visit prediction results for the medical institution.

[0064] In one possible implementation, the computer device performs k rounds of patient visit prediction processing based on the values ​​of the input parameters and k sets of random parameters to obtain k initial patient visit prediction results for the medical institution.

[0065] Step 204: Based on the initial predicted number of visits for k groups, determine the final predicted number of visits for the medical institution.

[0066] In possible implementations, a target operation is performed on the k initial predicted visitor counts to obtain the final predicted visitor counts for the medical institution. The target operation includes any one of the following: averaging, standard deviation, or variance.

[0067] Step 205: Based on the final predicted number of outpatient visits to the medical institution, determine the resource allocation recommendation information for the medical institution. The resource allocation recommendation information is used to provide suggestions when allocating resources to the medical institution.

[0068] In this embodiment, resources refer to the substances required by a user when seeking medical treatment at a medical institution. In possible implementations, resource allocation suggestion information includes equipment allocation suggestion information and / or physician allocation suggestion information. The equipment allocation suggestion information is used to provide suggestions when allocating equipment to the medical institution, and the physician allocation suggestion information is used to provide suggestions when allocating physicians to the medical institution.

[0069] This application embodiment can determine resource allocation suggestions for medical institutions based on the final predicted number of outpatient visits, thereby making the resource allocation of medical institutions more reasonable and maximizing the utilization efficiency of resources.

[0070] In summary, the technical solution provided in this application determines the values ​​of multiple sets of random parameters based on the correlation between input parameters and random parameters. Then, based on the input parameters and multiple sets of random parameters, multiple initial prediction results of patient visits are determined. Finally, based on the multiple initial prediction results of patient visits, the final prediction result of patient visits is determined. This application combines consideration of the correlation between various parameters and assigns values ​​to random parameters based on this correlation. By randomly simulating the distribution of possible result ranges (i.e., the aforementioned multiple sets of initial prediction results of patient visits), rather than just estimating results at a single point, the volatility of medium- and long-term predictions is considered, improving the accuracy of the final prediction result of patient visits. Based on the final prediction result of patient visits for the medical institution, resource allocation recommendations for the medical institution are determined. Since patient visits are fully considered during resource allocation, the resource allocation of the medical institution is more reasonable, maximizing resource utilization efficiency.

[0071] Please refer to Figure 3 The diagram illustrates a flowchart of a resource allocation method provided in another embodiment of this application. This method can be executed by a computer device and may include the following steps.

[0072] Step 301: Obtain input parameters related to the prediction of the number of visits to medical institutions.

[0073] In this embodiment of the application, the input parameters refer to user-input parameters related to the prediction of the number of visits to medical institutions.

[0074] Step 302: Determine the marginal distribution functions of the historical input parameters and historical random parameters based on the historical input parameters and historical random parameters, respectively.

[0075] The marginal distribution function of historical input parameters can also be called the marginal distribution function of historical input parameters; the marginal distribution function of historical random parameters can also be called the marginal distribution function of historical random parameters.

[0076] The marginal distribution function of historical input parameters refers to the probability distribution of historical input parameters when there is no knowledge of the marginal distributions of other historical input parameters and / or the marginal distribution functions of historical random parameters.

[0077] The marginal distribution function of historical random parameters refers to the probability distribution of historical random parameters when there is no knowledge of the marginal distributions of other historical random parameters and / or the marginal distribution functions of historical input parameters.

[0078] Step 303: Based on the marginal distribution function of historical input parameters and historical random parameters, the target connection function is obtained. The target connection function is used to indicate the target relevance.

[0079] Among possible implementations, the target connection function includes the Copula function. The Copula function, also known as a connection function, is a class of functions that connect the joint distribution function to their respective marginal distribution functions.

[0080] In possible implementations, the joint distribution between historical input parameters and historical random parameters is determined based on the marginal distribution functions of historical input parameters and historical random parameters; and the target connection function corresponding to historical input parameters and historical random parameters is determined based on the joint distribution between historical input parameters and historical random parameters.

[0081] In possible implementations, the joint distribution among historical random parameters is determined based on the marginal distribution function of each historical random parameter; and the target connection function corresponding to the historical random parameters is determined based on the joint distribution among the historical random parameters.

[0082] We map historical input parameters and historical random parameters to well-behaved distributions, which we can define as correlations. For example, these well-behaved distributions follow a normal distribution. For example, we map historical input parameters to a first normal distribution and historical random parameters to a second normal distribution; this mapping is a one-to-one mapping between quantiles. The joint distribution function of the first and second normal distributions is a bivariate normal distribution, based on which we can determine the joint distribution function of historical random parameters and historical input parameters, as well as the correlation structure.

[0083] For example, a Copula function is used to describe the correlation between two parameters. In this case, multiple Copula functions are needed to indicate the target correlation. For instance, assuming the historical input parameters include parameter 1 and parameter 2, and the historical random parameters include parameter 3 and parameter 4, then Copula function 1 is needed to indicate the correlation between parameter 1 and parameter 3, Copula function 2 is needed to indicate the correlation between parameter 1 and parameter 4, Copula function 3 is needed to indicate the correlation between parameter 2 and parameter 3, Copula function 4 is needed to indicate the correlation between parameter 2 and parameter 4, and Copula function 5 is needed to indicate the correlation between parameter 3 and parameter 4.

[0084] For example, a Copula function can be used to describe the correlation between multiple parameters. In this case, only one Copula function is needed to indicate the target correlation. For instance, assuming the historical input parameters include parameter 1 and parameter 2, and the historical random parameters include parameter 3 and parameter 4, then only Copula function 6 is needed to indicate the correlation between parameter 1 and parameter 3, the correlation between parameter 1 and parameter 4, the correlation between parameter 2 and parameter 3, the correlation between parameter 2 and parameter 4, and the correlation between parameter 3 and parameter 4.

[0085] Step 304: Determine k sets of random parameters based on the target connection function.

[0086] Once the target join function is determined, the correlation between the parameters is also determined. After the computer device obtains the input parameters, it can randomly generate k sets of random parameters based on the target join function and the input parameters.

[0087] Step 305: Based on the input parameters and k sets of random parameters, perform k rounds of patient visit prediction processing to obtain the initial patient visit prediction results for the medical institution.

[0088] In one possible implementation, the input parameters and k sets of random parameters are processed by a machine learning model to obtain k initial prediction results of the number of visits. The machine learning model refers to a neural network model used to predict the number of visits to medical institutions.

[0089] For example, input parameters and a set of random parameters are input into the machine learning model. The machine learning module processes the input parameters and random parameters to obtain a set of initial predictions for the number of patient visits. The above steps are repeated until the machine learning model has processed all k sets of random parameters and obtained k sets of initial predictions for the number of patient visits.

[0090] For example, the input parameters and k sets of random parameters are simultaneously input into the machine learning model to obtain k initial prediction results of the number of visits.

[0091] In a possible implementation, the input parameters include the total number of insured persons; the random parameters include the number of annual visits by individuals in the i-th age range who visit medical institutions of the j-th type, where i and j are positive integers; the computer device determines the initial prediction results of the number of visits in the following way: the total number of insured persons is divided into n categories of insured persons according to the n-th age range through the prediction model, where n is a positive integer greater than or equal to i; based on the number of annual visits by individuals and the number of persons in each category among the n-th insured persons, k groups of initial prediction results of the number of visits are determined.

[0092] In a possible implementation, since users will seek medical treatment at medical institutions designated by medical insurance, this embodiment of the application can determine the initial prediction of the number of visits based on the total number of insured persons. Because users of different age groups have different resource needs, in order to make the final resource allocation recommendation information more accurate, this embodiment of the application can determine the prediction of the number of visits based on age ranges, and then determine the resource allocation recommendation information based on the prediction of the number of visits.

[0093] Step 306: Based on the initial predicted number of visits for k groups, determine the final predicted number of visits for the medical institution.

[0094] Step 307: Based on the final predicted number of outpatient visits to the medical institution, determine the predicted resource demand information of the medical institution. The predicted resource demand information refers to the predicted resource demand of the medical institution.

[0095] In possible implementations, resource demand forecasting information includes equipment demand forecasting information and / or physician demand forecasting information. Equipment demand forecasting information refers to the predicted equipment demand of medical institutions, and physician demand forecasting information refers to the predicted physician demand of medical institutions.

[0096] In one example, resource demand forecasting information includes equipment demand forecasting information; in another example, resource demand forecasting information includes physician demand forecasting information; and in yet another example, resource demand forecasting information includes both equipment demand forecasting information and physician demand forecasting information.

[0097] Step 307 may include the following sub-steps:

[0098] First, determine the target number of patients who fall within the i-th age range and visit the j-th type of medical institution in the final predicted number of visits, where i and j are positive integers.

[0099] In possible implementations, the medical institution in this application embodiment refers to the j-th type of medical institution, and the final patient visit prediction result refers to the patient visit prediction result of the j-th type of medical institution.

[0100] Since age range and / or type of medical institution can affect resource demand, this embodiment of the application determines the target number of patient visits from the final predicted number of patient visits to determine resource demand prediction information.

[0101] Second, based on the predicted number of outpatient visits and the target correspondence, resource demand prediction information is determined. The target correspondence refers to the correspondence between users in the i-th age range who visit the j-th type of medical institution and their equipment needs, and / or the correspondence between users in the i-th age range who visit the j-th type of medical institution and their physician needs.

[0102] In one possible implementation, the computer device stores the target correspondence. After obtaining the target number of outpatient visits prediction results, the computer device can determine the resource demand prediction information based on the target number of outpatient visits prediction results and the target correspondence.

[0103] Step 308: Based on the resource demand forecast information of medical institutions, determine the resource allocation recommendations for medical institutions.

[0104] In one possible implementation, once the computer equipment has determined the equipment demand forecast information, it can determine equipment allocation recommendations for the medical institution based on that information. These recommendations are used to suggest equipment allocations to the medical institution. For example, the equipment allocation recommendations could be to allocate the number of devices to the medical institution that correspond to the equipment demand forecast information.

[0105] In one possible implementation, once the computer device determines the physician demand forecast information, it can determine physician allocation recommendations for medical institutions based on this forecast information. These recommendations are used to suggest which physicians to allocate to the medical institution. For example, the physician allocation recommendations could be the number of physicians to be allocated to the medical institution corresponding to the physician demand forecast information.

[0106] It should be noted that the embodiments of this application only use equipment and physicians as examples for illustration. In other possible implementations, resources can also be other types of material elements, and the embodiments of this application do not limit this.

[0107] In an illustrative embodiment, after the computer device performs k rounds of visitor prediction processing based on input parameters and k sets of random parameters to obtain k initial visitor prediction results for the medical institution, the following process can also be executed:

[0108] First, based on the initial predicted number of outpatient visits for group k, determine the initial predicted increase in medical insurance fund and the initial predicted consumption of medical insurance fund for group k.

[0109] In possible implementations, the input parameters are related not only to the predicted number of outpatient visits to medical institutions but also to the increase and depletion of the medical insurance fund. For example, the input parameters may include at least one of the following: total number of insured persons, employed ratio, average wage, dependency ratio, average annual number of illnesses per person, average cost per outpatient visit, number of outpatient visits, and number of people in different age groups. Total number of insured persons includes any one of the following: total number of urban employees, total number of insured persons for urban and rural residents, or total number of insured persons for both urban employees and urban and rural residents. Employed ratio refers to the proportion of employed insured persons to the total number of insured persons. Average wage indicates the overall wage level. Dependency ratio, also known as the population dependency ratio, indicates the ratio of the non-working-age population to the working-age population, illustrating approximately how many non-working-age people are supported by every 100 working-age people. Average annual number of illnesses per person indicates the average number of illnesses per person per year. Average cost per outpatient visit indicates the average cost per person per outpatient visit. Number of outpatient visits indicates the number of outpatient visits. Number of people in different age groups indicates the degree of population aging.

[0110] In a possible implementation, the initial predicted increases and decreases of the medical insurance fund (k groups) can be determined using the aforementioned machine learning model. The machine learning model processes the input parameters and k groups of random parameters to obtain the initial predicted number of outpatient visits (k groups). Then, the machine learning model further processes the input parameters, random parameters, and the initial predicted number of outpatient visits (k groups) to obtain the initial predicted increases and decreases of the medical insurance fund (k groups).

[0111] In possible implementations, the predicted increases and decreases of the initial medical insurance fund for k groups can be obtained through a prediction model. The prediction model is based on relevant policies. Since the relevant policies for urban employees and rural residents are not entirely consistent, to more accurately predict the increase and decrease of the medical insurance fund, the prediction model is divided into two parts: one for urban employees and one for rural residents. This decomposes the increase and decrease of the medical insurance fund at both ends, providing support for macroeconomic decision-making. In possible implementations, the prediction model includes any one of the following: a model predicting the increase and decrease of the medical insurance fund for urban employees; a model predicting the increase and decrease of the medical insurance fund for rural residents; or a model predicting the increase and decrease of the medical insurance fund for both urban employees and rural residents. This application does not limit the specific model to these aspects. When the prediction model includes a model predicting the increase and depletion of the medical insurance fund for urban employees, the input parameters and random parameters are correlated with the predicted increase and depletion of the medical insurance fund for urban employees. When the prediction model includes a model predicting the increase and depletion of the medical insurance fund for urban and rural residents, the input parameters and random parameters are correlated with the predicted increase and depletion of the medical insurance fund for urban and rural residents. When the prediction model includes models predicting the increase and depletion of the medical insurance funds for both urban employees and urban and rural residents, the input parameters and random parameters are correlated with the predicted increase and depletion of the medical insurance fund for both urban employees and urban and rural residents. In possible implementations, the prediction model construction process considers the distribution and constraints of different parameter adaptations, fully simulates relevant policies such as deductibles, reimbursement ratios, and annual limits, and aggregates data upwards from each insured person's each medical visit as the smallest unit.

[0112] Among possible implementations, the prediction model employs a DFA model, such as... Figure 4The diagram illustrates the framework of the DFA model. The construction process of the DFA model considers the distribution and constraints of different parameter adaptations, fully simulating relevant policies such as deductibles, reimbursement ratios, and annual limits, aggregating upwards from each insured person's individual medical visit as the smallest unit. In each simulation, the DFA model considers both the increase and depletion of the medical insurance fund, and is divided into two independent business modules: urban employees and rural residents. A random scenario generator determines the number of simulations, k. Each simulation represents a possible outcome of the current medical insurance operation; that is, each simulation represents an initial prediction of the increase and depletion of the medical insurance fund. For example, k can be 10000. During each simulation, the DFA model randomly generates a set of parameter values ​​to predict the initial increase and depletion of the medical insurance fund. The inputs to the DFA model include historical data and model parameters. Historical data refers to historical increases and depletions of the medical insurance fund, and model parameters include input parameters and random parameters. Strategic assumptions refer to the combination of policies related to the prediction of the increase and consumption of medical insurance funds. The DFA model outputs the prediction results of the increase and consumption of medical insurance funds, and users can analyze and output adjustment strategies based on the prediction results of the increase and consumption of medical insurance funds.

[0113] Different factors influencing the increase and depletion of the medical insurance fund exhibit varying degrees of correlation, particularly tail correlation. Dependence also exists between the increase and depletion of the medical insurance fund; if tail correlation exists, the parameters are non-linearly correlated. Generally, the increase and depletion of the medical insurance fund are considered relatively independent. However, in extreme cases, a large population's demand for services and medical care may lead to a significant negative correlation between the increase and depletion of the medical insurance fund, thereby exacerbating the risk of fund depletion. Therefore, in practical applications, this embodiment introduces target correlation into the prediction model, utilizing the correlation between random parameters and / or the correlation between random parameters and input parameters to generate relevant random numbers, which are then assigned values ​​to the random parameters of the prediction model. This embodiment considers sudden public health events, improving the accuracy of the final predicted increases and depletion results of the medical insurance fund.

[0114] In an illustrative embodiment, the following examples will use the following as an example to illustrate the data processing flow: input parameters include the total number of insured persons; random parameters include the number of annual outpatient visits per person in the i-th age range and the cost of a single outpatient visit per person in the j-th age range and the cost of a single outpatient visit per person in the i-th age range and the j-th age range. i and j are positive integers. In a possible implementation, the computer device determines k sets of initial predicted increases in medical insurance funds and k sets of initial predicted consumption of medical insurance funds in the following manner: the total number of insured persons is divided into n categories of insured persons according to n age ranges using a prediction model; the actual amount of medical insurance contributions is determined based on the corresponding medical insurance contribution policies for each of the n categories of insured persons; the initial predicted increase in medical insurance funds is determined based on the actual amount of medical insurance contributions; the annual reimbursement amount for each person is determined based on the k sets of initial predicted outpatient visits, the cost of a single outpatient visit, and the corresponding medical insurance reimbursement policies for each of the n categories of insured persons; and the initial predicted consumption of medical insurance funds is determined based on the annual reimbursement amount for each person. For example, the number of annual outpatient visits per person follows a Poisson distribution. For example, the cost of a single medical visit follows a gamma distribution.

[0115] For example, the annual number of outpatient visits for an individual in the i-th age range who seeks treatment at a j-th type of medical institution includes at least one of the following: the annual number of hospitalizations for an individual in the i-th age range who is hospitalized at a j-th type of medical institution, and the annual number of outpatient visits for an individual in the i-th age range who is outpatient at a j-th type of medical institution.

[0116] For example, the cost of a single visit to a medical institution of type j for someone in the i-th age range includes at least one of the following: the cost of a single hospitalization for someone in the i-th age range who is hospitalized at a medical institution of type j, and the cost of a single outpatient visit for someone in the i-th age range who is outpatient at a medical institution of type j.

[0117] In one example, the total number of insured persons includes the total number of insured urban employees. In this case, the predictive model can divide the total number of insured urban employees into n categories of insured persons based on the urban employees' medical insurance payment policy and the n age ranges.

[0118] Taking the relevant medical insurance payment policy for urban employees as an example, the insured urban employees include all employers within the city's administrative region, including enterprises, government agencies, public institutions, social organizations, and private non-enterprise units (hereinafter referred to as employers) and their employees and retirees. The total number of insured persons includes the number of employed insured persons and the number of retired insured persons. Let B represent the total number of insured persons, assuming that B follows a normal distribution with a lower bound of 0. Let I represent the proportion of employed insured persons in the total number of insured persons, assuming that I follows a beta distribution, then the number of employed insured persons B S For: B s =B×M; Number of retired insured persons B D For: B D=B×(1-M).

[0119] To obtain a more accurate number of insured individuals, the total number of insured individuals is categorized according to different age groups. For example, the predictive model divides the total number of insured individuals into five age ranges: under 35, 35 to 45, over 45, under 70, and over 70.

[0120] Assume that the proportion of employed insured persons in different age groups in the total number of insured persons is as shown in Table 1 below:

[0121] Table 1

[0122] age group Under 35 years old 35 to 45 years old 45 years and older Proportion a b c

[0123] Where a represents the proportion of employed insured persons under the age of 35 in the total number of employed insured persons, b represents the proportion of employed insured persons aged 35 to 45 in the total number of employed insured persons, and c represents the proportion of employed insured persons over the age of 45 in the total number of employed insured persons, and a+b+c=1.

[0124] Assume that the proportion of retirees in different age groups in the total number of retirees is as shown in Table 2 below:

[0125] Table 2

[0126] age group Under 70 years old 70 years and older Proportion d e

[0127] Where d represents the proportion of retirees under 70 years old in the total number of retirees, e represents the proportion of retirees over 70 years old in the total number of retirees, and d+e=1.

[0128] Let i represent the type of insured person, then:

[0129]

[0130] The five categories of insured individuals are denoted as B. i (i = 1, 2, 3, 4, 5), we have:

[0131]

[0132] Taking the relevant medical insurance payment policy for urban employees as an example, the medical insurance premium payment for urban employees includes two parts: individual payment and employer payment. Let F represent the total amount of medical insurance paid by urban employees, D represent the total amount of individual medical insurance payment, and E represent the total amount of employer medical insurance payment, then we have: F = D + E.

[0133] Let C represent the total individual contribution base. According to the relevant medical insurance contribution policy for urban employees, the individual contribution amount is 2% of the individual contribution base plus 3 yuan. Therefore: D = 2% × C + 3 × B × M;

[0134] E represents 10% of the total employer contribution base, where the employer contribution base is the sum of the individual contribution bases of all employees within the company. The total employer contribution base equals the total individual contribution base using the following formula:

[0135]

[0136] Therefore, E = 10% × C.

[0137] The relevant medical insurance payment policies for urban employees stipulate the following regarding the individual contribution base: If an employee's average monthly salary in the previous year is lower than 60% of the average monthly salary of employees in the city in the previous year, then 60% of the average monthly salary of employees in the city in the previous year shall be used as the contribution wage base for basic medical insurance premiums; if an employee's average monthly salary in the previous year exceeds 300% of the average monthly salary of employees in the city in the previous year, the portion exceeding 300% shall not be used as the contribution wage base, and no basic medical insurance premiums shall be paid. When an employee has accumulated 25 years of basic medical insurance premiums for men and 20 years for women, and has completed retirement procedures according to national regulations, receiving a monthly basic pension or retirement pay, they enjoy the basic medical insurance benefits of retirees and are no longer required to pay basic medical insurance premiums. Using the above example, employed insured individuals are divided into three categories, with their individual salaries denoted as A, B, C, and D respectively. i (i = 1, 2, 3), based on the relevant medical insurance payment policies for urban employees mentioned above, we can assume A i The gamma distribution follows a truncated gamma distribution with parameters that are not necessarily the same, but with the same upper and lower bounds. Let avginc represent the relevant social average wage, then the upper bound of the truncated gamma distribution is: upper = 300% × avginc; the lower bound of the truncated gamma distribution is: lower = 60% × avginc.

[0138] The total individual contribution base for the three types of employed insured individuals is denoted as C. i (i = 1, 2, 3), we have:

[0139] Since retirees do not pay medical insurance premiums, the total individual contribution base C can be determined by the following formula: C = C1 + C2 + C3, where C1 represents the total individual contribution base of the first category of employed insured persons, C2 represents the total individual contribution base of the second category of employed insured persons, and C3 represents the total individual contribution base of the third category of employed insured persons.

[0140] In practice, not all employed insured individuals actually pay their premiums; therefore, the collection rate needs to be considered.i (i = 1, 2, 3) represent the contribution rates for the three types of employed insured individuals, and U4 represents the contribution rate for the employer. Therefore, the total actual individual contribution to medical insurance is calculated as follows. for:

[0141] Actual total amount of medical insurance paid by the employer for:

[0142] Therefore, the actual amount of medical insurance paid by urban employees for:

[0143] In another example, the total number of insured persons includes the total number of urban and rural residents enrolled in medical insurance. In this case, the predictive model can divide the total number of urban and rural residents enrolled in medical insurance into n categories of insured persons based on the medical insurance payment policies for urban and rural residents.

[0144] Taking the relevant medical insurance payment policy for urban and rural residents as an example, the participants in the basic medical insurance for urban and rural residents are the following three categories of people who have no other basic medical security: 1. Urban and rural residents with local household registration who are male and over 60 years old or female and over 50 years old (hereinafter referred to as urban and rural elderly); 2. Urban and rural residents with local household registration who are male and under 60 years old or female and under 16 years old (hereinafter referred to as residents of working age); 3. Students with local household registration who are enrolled in full-time regular institutions of higher learning (including private institutions of higher learning), research institutes, ordinary primary and secondary schools, secondary vocational schools, special education schools, and work-study schools within the administrative region of the city, as well as local residents under the age of 16 who are not enrolled in school; and full-time non-working non-Beijing students who are receiving regular higher education in full-time regular institutions of higher learning (including private institutions of higher learning) and research institutes within the administrative region of the city (hereinafter referred to as students and children).

[0145] For example, the total number of insured persons is divided into three categories based on three age ranges using a predictive model. The three age ranges are: men aged 60 and above and women aged 50 and above; men aged 16 and above but under 60 and women aged 16 and below 50; and men under 16 and women under 16.

[0146] Let i represent the type of insured person, then:

[0147]

[0148] The number of participants in the three categories is denoted as Q. i (i = 1, 2, 3), assume Q i They follow a normal distribution with parameters that are not necessarily the same, but all of them have a lower limit of 0.

[0149] Taking the relevant medical insurance payment policy for urban and rural residents as an example, the medical insurance premium payment for urban and rural residents includes two parts: personal payment and government subsidy. Let FJ represent the total urban and rural residents' medical insurance funding, FJ1 represent the total amount of personal medical insurance payment, and FJ2 represent the total amount of government subsidy for medical insurance, then we have: FJ = FJ1 + FJ2;

[0150] According to the relevant medical insurance payment policies for urban and rural residents, the individual medical insurance payment policy stipulates that the elderly and students / children pay 300 yuan per person per year, while residents of working age pay 520 yuan per person per year. Therefore, the total individual medical insurance payment FJ1 can be determined using the following formula:

[0151] FJ1 = 300 × (Q1 + Q2) + 520 × Q3;

[0152] According to the relevant government subsidy policies for urban and rural residents' medical insurance, the elderly receive a subsidy of 4,180 yuan per person per year, of which 1,860 yuan is from the municipal government and 2,320 yuan is from the district government; students and children receive a subsidy of 1,610 yuan per person per year, of which 575 yuan is from the municipal government and 1,035 yuan is from the district government; and residents of working age receive a subsidy of 2,150 yuan per person per year, of which 845 yuan is from the municipal government and 1,305 yuan is from the district government. Therefore, the total government subsidy for medical insurance, FJ2, can be determined using the following formula:

[0153] FJ2=4180×Q1+1610×Q2+2150×Q3;

[0154] In practice, not all urban and rural residents enrolled in the insurance program actually pay their premiums; therefore, the collection rate needs to be considered. i (i = 1, 2, 3) represent the contribution rates for the three categories of urban and rural residents enrolled in medical insurance, respectively. Therefore, the actual amount of medical insurance paid by urban and rural residents is... It can be determined using the following formula:

[0155]

[0156] Therefore, the actual total funding for urban and rural residents' medical insurance for:

[0157] In one possible implementation, when the total number of insured persons includes the total number of insured urban employees, the actual amount of medical insurance paid by urban employees is used as the initial predicted increase in the medical insurance fund.

[0158] In one possible implementation, when the total number of insured persons includes the total number of urban and rural residents enrolled in medical insurance, the initial predicted increase in the medical insurance fund is determined based on the actual amount of medical insurance contributions paid by urban and rural residents and the amount of medical insurance subsidies provided by the government. For example, the sum of the actual amount of medical insurance contributions paid by urban and rural residents and the amount of medical insurance subsidies provided by the government can be used as the initial predicted increase in the medical insurance fund.

[0159] In one possible implementation, when the total number of insured persons includes both urban employees and rural residents, the initial predicted increase in the medical insurance fund is determined based on the actual medical insurance contributions paid by urban employees, the actual medical insurance contributions paid by rural residents, and the government-subsidized medical insurance amount. For example, the sum of the actual medical insurance contributions paid by urban employees, the actual medical insurance contributions paid by rural residents, and the government-subsidized medical insurance amount can be used as the initial predicted increase in the medical insurance fund.

[0160] In possible implementations, an individual's annual number of outpatient visits includes the number of outpatient visits, the number of hospitalizations, and the number of major illnesses; the cost per visit includes the cost per outpatient visit (also known as the average cost per outpatient visit), the cost per hospitalization (also known as the average reimbursement amount per hospitalization visit), and the cost per major illness; the individual's annual reimbursement includes the annual reimbursement for outpatient visits, the annual reimbursement for hospitalizations, and the annual reimbursement for major illnesses. For example, the annual outpatient reimbursement is determined based on the number of outpatient visits, the cost per outpatient visit, and the corresponding medical insurance reimbursement policies for each of the n categories of insured individuals; the annual hospitalization reimbursement is determined based on the number of hospitalizations, the cost per hospitalization, and the corresponding medical insurance reimbursement policies for each of the n categories of insured individuals; and the annual major illness reimbursement is determined based on the number of major illnesses, the cost per major illness, and the corresponding medical insurance reimbursement policies for each of the n categories of insured individuals.

[0161] Among the possible implementation methods, the annual outpatient reimbursement limit for urban employees can be determined in the following ways:

[0162] Taking the relevant medical insurance reimbursement policy for urban employees as an example, outpatient medical expenses fall within the scope of payment from the pooled account. Since relevant data on outpatient medical expenses is difficult to obtain, from an actuarial perspective, when predicting outpatient medical expenses, two variables can be considered: the annual number of outpatient visits and the average reimbursement amount per outpatient visit. Some core parameter variables were used in the prediction process.

[0163] Let G ij Let G represent the number of outpatient visits per year for individuals within the i-th age range who visit medical institutions of the j-th type. ij Follows a Poisson distribution, when G ij When the value is 0, the person has no outpatient experience within one year.

[0164] Let J1 represent the cost of a single outpatient visit, and assume that J1 follows a gamma distribution. If we consider that J1 behaves differently in different age ranges and different types of medical institutions, then let J1 ij represent the cost of a single outpatient visit for a person in the i-th age range and receiving outpatient treatment at the j-th type of medical institution, and assume that J1 ij follows a gamma distribution with parameters that may not be the same.

[0165] Combining the annual number of outpatient visits per individual and the cost of a single outpatient visit, we obtain the total annual outpatient cost for an individual as follows:

[0166] The urban employee medical insurance outpatient reimbursement policy included in the relevant medical insurance reimbursement policies for urban employees is shown in Table 3 below:

[0167] Table 3

[0168]

[0169] Let J1qpe ij , J1bxb ij , J1fd ij respectively represent the deductible amount, reimbursement ratio, and cap amount of the total annual outpatient cost for a person in the i-th age range and receiving outpatient treatment at the j-th type of medical institution. Then, the annual outpatient reimbursement amount from the medical insurance pooling account for a person in the i-th age range and receiving outpatient treatment at the j-th type of medical institution is as follows:

[0170]

[0171] In a possible implementation, the annual inpatient reimbursement amount for urban employees can be determined as follows: <

[0172] Let H ij represent the annual number of inpatient visits for a person in the i-th age range and receiving inpatient treatment at the j-th type of medical institution. Assume that H ij follows a Poisson distribution. When H ij = 0, this person has no inpatient experience within a year.

[0173] In an actual scenario, H ij is also restricted by the hospital beds and thus cannot be infinitely large. Let cw j represent the actual number of beds in the j-th type of hospital, and assume there are 365 days in a year. Then H ij has an upper limit:

[0174] If we consider whether the inpatient is a first-time inpatient, then let P ij ]Let P represent the probability that an inpatient in the i-th age range who is treated at a j-th type of medical institution is admitted to the hospital for the first time in that year. ij Following a beta distribution with not necessarily identical parameters, the number of first hospitalizations in that year for individuals within the i-th age range who visited a j-th type of medical institution is H1. ij For: H1 ij =I[I(H ij ≠0)×P ij >0.5]; where I(*) is an indicator function, which takes the value of 1 when the condition is met, and 0 otherwise; H2 is the number of non-first-time hospitalizations in the i-th age range and at the j-th type of medical institution. ij H2 ij =H ij -H1 ij .

[0175] Let J2 denote the cost of a single hospitalization. Since J2 is always non-negative, assume that J2 follows a gamma distribution.

[0176] If we consider that J2's performance varies across different age ranges and types of healthcare institutions, then let J2 ij Let J represent the cost of a single hospitalization for individuals within the i-th age range who are hospitalized at a medical institution of type j, and assume J2 ij They follow a gamma distribution with not necessarily the same parameters.

[0177] The relevant medical insurance reimbursement policies for urban employees, including the inpatient reimbursement policies, are shown in Table 4 below:

[0178] Table 4

[0179]

[0180] Let k denote the cost interval, then:

[0181] Let J2qpe ijk1 J2bxb ijk1 J2fd ijk1 Let represent the deductible, reimbursement ratio, and maximum reimbursement limit for a single hospitalization for individuals within the i-th age range who receive treatment at a j-th type of medical institution, and for those within the k-th cost range who are hospitalized for the first time. Then, the medical insurance pooled account will reimburse the individual for a single hospitalization within the i-th age range who receive treatment at a j-th type of medical institution, and for those within the k-th cost range who are hospitalized for the first time. for:

[0182] Let J2qpe ijk2 J2bxb ijk2 J2fd ijk2Let represent the deductible, reimbursement ratio, and maximum reimbursement limit for a single hospitalization for individuals within the i-th age range who receive treatment at a j-th type of medical institution, and for those within the k-th cost range who are not first-time hospitalized. Then, the medical insurance pooled account will reimburse the individual for a single hospitalization for those within the i-th age range who receive treatment at a j-th type of medical institution, and for those within the k-th cost range who are not first-time hospitalized. for:

[0183] The individual's annual hospitalization reimbursement amount is calculated by combining the number of hospitalizations per year with the reimbursement amount for each hospitalization. for:

[0184] Among the possible implementation methods, the annual reimbursement amount for serious illnesses for urban employees can be determined in the following ways:

[0185] Taking the relevant medical insurance reimbursement policy for urban employees as an example, for out-of-pocket medical expenses exceeding the deductible of 39,525 yuan, up to 50,000 yuan (inclusive) in total, 60% will be covered by the Urban Employee Major Medical Mutual Aid Fund; for out-of-pocket medical expenses exceeding 50,000 yuan (exclusive), 70% will be covered by the Urban Employee Major Medical Mutual Aid Fund, with no upper limit. Urban employee major illness medical insurance is settled once a year.

[0186] Annual outpatient co-payment for individuals in age range i who are hospitalized at medical institutions of type j for:

[0187] Annual outpatient co-payment for individuals in age range i who are hospitalized at medical institutions of type j for:

[0188] Annual out-of-pocket medical expenses for individuals in age range i who are hospitalized in medical institutions of type j for:

[0189] Let dbbz, dbqj, dbbxb1, and dbbxb2 represent the deductible for serious illnesses, the range of reimbursement rates for the first tier of deductibles for super serious illnesses, the first tier reimbursement rate for serious illnesses, and the second tier reimbursement rate for serious illnesses, respectively. Then, the annual reimbursement amount DB for serious illnesses from the Major Medical Mutual Aid Fund for individuals within the i-th age range and at medical institutions of the j-th type is... ij for:

[0190] Among the possible implementation methods, the annual outpatient reimbursement amount for urban and rural residents can be determined in the following ways:

[0191] G2 ijThis represents the number of outpatient visits per year for individuals within the i-th age range who visit medical institutions of the j-th type. Let G2 be an example. ij Follows a Poisson distribution, when G2 ij When the value is 0, it means that the person has not had any outpatient experience within the past year.

[0192] Let M1 represent the cost of a single outpatient visit, and assume that M1 follows a gamma distribution. If we consider that M1 varies across different age groups and types of healthcare institutions, then let M1... ij Let M1 represent the cost of a single outpatient visit for individuals within the i-th age range who seek medical care at a j-th type of medical institution, and assume M1 ij They follow a gamma distribution with not necessarily the same parameters.

[0193] The total annual outpatient cost is calculated by combining the number of outpatient visits per individual per year with the total annual outpatient cost per individual per year. for:

[0194] The relevant medical insurance reimbursement policies for urban and rural residents, including the outpatient reimbursement policies, are shown in Table 5 below:

[0195] Table 5

[0196]

[0197] Let M1qpe ij M1bxb ij M1fd ij Let represent the deductible, reimbursement ratio, and maximum reimbursement limit for outpatient expenses of individuals within the i-th age range who receive treatment at type j medical institutions, respectively. Then, the annual outpatient reimbursement amount from the medical insurance pooled account for individuals within the i-th age range who receive treatment at type j medical institutions is... for:

[0198]

[0199] Among the possible implementation methods, the annual inpatient reimbursement amount for urban and rural residents can be determined in the following ways:

[0200] Let H21 ij This represents the number of hospitalizations per year for an individual within the i-th age range who visits a medical institution of the j-th type, assuming H21 ij Follows a Poisson distribution, when H21 ij When the value is 0, the person has not been hospitalized within the past year.

[0201] In practice, H21 ij It is also limited by the number of hospital beds, and cannot be infinitely large. (This is followed by an unrelated phrase: "Let CW") j Let H21 represent the actual number of beds in the j-th type of medical institution, and assuming there are 365 days in a year, then H21ij There is an upper limit:

[0202] If we consider whether the inpatient is a first-time hospitalization, then let P2 ij Let P2 represent the probability that an inpatient in the i-th age range and at the j-th type of medical institution is admitted for the first time in that year. ij Following a beta distribution with not necessarily identical parameters, the number of first hospitalizations in that year for individuals within the i-th age range who visited a j-th type of medical institution is H21. ij For: H21 ij =I[I(H2) ij ≠0)×P2 ij >0.5]. Where I(*) is the indicator function, which takes a value of 1 when the condition is met, and 0 otherwise. The number of non-first-time hospitalizations H22 within the i-th age range and at the j-th type of medical institution. ij H22 ij =H2 ij -H21 ij .

[0203] Let M2 be the average cost per hospitalization for urban and rural residents, and assume that M2 follows a gamma distribution. If we consider that M2 varies across different age groups and types of medical institutions, then M2... ij Let M2 represent the cost of a single hospitalization for individuals within the i-th age range who are hospitalized at a j-th type of medical institution, and assume M2 ij They follow a gamma distribution with not necessarily the same parameters.

[0204] The relevant medical insurance reimbursement policies for urban and rural residents, including the inpatient reimbursement policies, are shown in Table 6 below:

[0205] Table 6

[0206]

[0207] Let M2qpe ij1 M2bxb ij1 M2fd ij1 Let represent the deductible, reimbursement ratio, and maximum reimbursement limit for a single hospitalization for an individual within the i-th age range and at a j-th type of medical institution for their first hospitalization. Then, the medical insurance pooled account will reimburse the individual for a single hospitalization within the i-th age range and at a j-th type of medical institution for their first hospitalization. for:

[0208]

[0209] Let M2qpeij2, M2bxbij2, and M2fdij2 represent the deductible, reimbursement ratio, and maximum reimbursement limit for a single hospitalization for individuals within the i-th age range and at a j-th type of medical institution who are not receiving their first hospitalization. Then, the medical insurance pooled account will reimburse the individual for a single hospitalization within the i-th age range and at a j-th type of medical institution who are not receiving their first hospitalization. for:

[0210]

[0211] The individual's annual hospitalization reimbursement amount is calculated by combining the number of hospitalizations per year with the reimbursement amount for each hospitalization. for:

[0212]

[0213] Among the possible implementation methods, the annual reimbursement amount for serious illnesses for urban and rural residents can be determined in the following ways:

[0214] Taking the relevant medical insurance reimbursement policies for urban and rural residents as an example, after basic medical insurance reimbursement, if an insured resident's out-of-pocket expenses within the scope of basic medical insurance exceed the per capita disposable income of the lowest 20% of urban residents in the city in the previous year, they can receive a subsidy from the urban and rural residents' major illness medical insurance. In 2019, the deductible for urban and rural residents' major illness insurance was 30,404 yuan. Regarding the reimbursement ratio, for out-of-pocket medical expenses exceeding the deductible (excluding the deductible) up to 50,000 yuan, the major illness insurance fund pays 65%; for out-of-pocket medical expenses exceeding 50,000 yuan (excluding the deductible), the major illness insurance fund pays 75%, with no upper limit.

[0215] Annual outpatient co-payment for individuals in age category i who are hospitalized at type j medical institutions for:

[0216] Annual outpatient co-payment for individuals in age category i who are hospitalized at type j medical institutions for:

[0217] Annual out-of-pocket medical expenses for individuals in age range i who are hospitalized in medical institutions of type j for:

[0218] Let dbbzj, dbqjj, dbbxb1j, and dbbxb2j represent the deductible for serious illnesses, the range of reimbursement rates for first-tier deductibles for super-serious illnesses, the first-tier reimbursement rate for serious illnesses, and the second-tier reimbursement rate for serious illnesses, respectively. Then, the annual reimbursement amount for serious illnesses from the Major Medical Mutual Aid Fund for individuals of age i and hospital type j is DBJ. ij for:

[0219]

[0220] Among the possible implementation methods, the overall annual medical reimbursement amount is determined based on the individual's annual medical reimbursement amount, the number of people in each category among the n types of insured persons, the number of medical institution types, and the initial prediction results of the number of medical visits; and the initial prediction results of the medical insurance fund consumption are determined based on the overall annual medical reimbursement amount.

[0221] Among the possible implementation methods, the total annual medical reimbursement amount includes the total annual outpatient reimbursement amount, the total annual inpatient reimbursement amount, and the total annual major illness reimbursement amount.

[0222] In possible implementations, the overall annual outpatient reimbursement amount for urban employees can be determined as follows: From an actuarial perspective, a composite model is used to predict the overall annual outpatient reimbursement amount L1 for urban employees, resulting in:

[0223] In possible implementations, the overall annual inpatient reimbursement amount for urban employees can be determined as follows: Using a composite model to calculate the overall annual inpatient reimbursement amount L2, we have:

[0224] In possible implementations, the overall annual reimbursement amount for serious illnesses for urban employees can be determined as follows: Using a composite model to calculate the overall annual reimbursement amount (DB) for serious illnesses, we have:

[0225] In possible implementation methods, the overall annual outpatient reimbursement amount for urban and rural residents can be determined as follows: From an actuarial perspective, a composite model is used to calculate the overall annual outpatient reimbursement amount L5, resulting in:

[0226] In possible implementations, the overall annual inpatient reimbursement amount for urban and rural residents can be determined as follows: Using a composite model to calculate the overall annual inpatient reimbursement amount L4, we have:

[0227] Among possible implementation methods, the overall annual reimbursement amount for major illnesses for urban and rural residents can be determined as follows: Using a composite model to calculate the total inpatient medical expenses (DBJ), we have:

[0228] It should be noted that the above description is merely exemplary and may be adjusted according to different relevant policies. This application embodiment does not limit this.

[0229] In possible implementations, when the total number of insured persons includes the total number of insured urban employees, the initial predicted medical insurance fund consumption is also related to the balance of the individual accounts of urban employees. The balance of the individual accounts can be determined in the following way:

[0230] The relevant medical insurance policies for urban employees establish personal accounts for insured individuals. These accounts consist of basic medical insurance premiums paid by the employee, basic medical insurance premiums paid by the employer and transferred to the personal account according to regulations, interest on the account balance, and other funds legally included in the personal account. Funds in the personal account are solely for personal medical payments and cannot be used to pay for medical insurance fund expenditures.

[0231] The percentage of premiums paid by employers and transferred to individual accounts is shown in Table 7 below:

[0232] Table 7

[0233]

[0234] Therefore, the total amount of funds in the individual account, L3, is:

[0235]

[0236] For example, the initial predicted medical insurance fund consumption for urban employees is determined based on the total annual inpatient reimbursement amount, the total annual outpatient reimbursement amount, the total annual major illness reimbursement amount, and the balance of their personal accounts. For instance, the sum of the total annual inpatient reimbursement amount, the total annual outpatient reimbursement amount, the total annual major illness reimbursement amount, and the balance of their personal accounts is determined as the initial predicted medical insurance fund consumption for urban employees.

[0237] For example, the initial predicted medical insurance fund consumption for urban and rural residents is determined based on the total annual inpatient reimbursement, total annual outpatient reimbursement, and total annual catastrophic illness reimbursement. For instance, the sum of the total annual inpatient reimbursement, total annual outpatient reimbursement, and total annual catastrophic illness reimbursement is used as the initial predicted medical insurance fund consumption for urban and rural residents.

[0238] This application's embodiments calculate the annual number of outpatient visits from the total number of insured individuals, as well as the number of visits per visit and the average total medical cost per visit for each individual. This application's embodiments consider influencing factors such as the visitor's age and the level of the medical institution while predicting the average medical cost per visit; by defining the scope of medical insurance coverage for the total cost and incorporating policy-related factors, the accuracy of the final predicted increase and depletion of the medical insurance fund is improved.

[0239] Second, based on the initial predicted increases and decreases of the medical insurance fund in k groups, the final predicted increases and decreases of the medical insurance fund are determined.

[0240] In possible implementations, a target operation is performed on k initial predicted increases in medical insurance funds to obtain the final predicted increase in medical insurance funds. The target operation includes any one of the following: average calculation, mean square error calculation, or variance calculation.

[0241] In possible implementations, a target operation is performed on k initial medical insurance fund consumption prediction results to obtain the final medical insurance fund consumption prediction result. The target operation includes any one of the following: average calculation, mean square error calculation, and variance calculation.

[0242] Third, based on the final predicted increase and the final predicted depletion of the medical insurance fund, the final predicted balance of the medical insurance fund is determined.

[0243] For example, the difference between the final predicted increase in medical insurance funds and the final predicted depletion of medical insurance funds is determined as the final predicted balance of medical insurance funds. In possible implementations, the predicted balance of medical insurance funds includes the predicted balance of urban employee medical insurance funds, the predicted balance of urban and rural resident medical insurance funds, and the sum of the predicted balances of urban employee medical insurance funds and urban and rural resident medical insurance funds.

[0244] Fourth, determine the value at risk of the final predicted balance of the medical insurance fund under a pre-set confidence level. The value at risk is used as a threshold to indicate the degree of loss.

[0245] In one of the possible implementations, computer devices can determine the value at risk in the following ways:

[0246] 1. Based on the pre-set confidence level, determine the initial value at risk (VaR) of the loss amount of the final predicted medical insurance fund balance. The VaR is used to indicate the maximum possible loss of the final predicted medical insurance fund balance.

[0247] The preset reliability is determined by the user; for example, the preset reliability can be 95%, 97.5%, or 99%.

[0248] In one of the possible implementations, the initial value at risk VaRα(X) can be determined by the following formula:

[0249] VaR α (X)=inf{x|Pr(X≤x)≥α};

[0250] Where X represents the predicted loss amount of the final medical insurance fund balance, and α represents the pre-set confidence level.

[0251] 2. Obtain the distribution density of the loss amount in the final predicted result of the medical insurance fund balance.

[0252] 3. Determine the value at risk based on the distribution density of the loss amount, the initial value at risk, and the pre-set confidence level of the final predicted balance of the medical insurance fund.

[0253] In its possible implementation, the Value at Risk (TailVaR) can be determined using the following formula:

[0254]

[0255] Among them, f x (x) is the probability density function of the random variable X.

[0256] In this embodiment, the accuracy of the predicted medical insurance fund balance can be improved by using TVaR (Tail Value at Risk). TailVaR is a new risk measurement method that meets the consistency requirement. It is expressed as an expected value, referring to the expected value of loss exceeding VaRα(X), i.e.: TailVaR = E[X|X≥VaR] α (X)].

[0257] In possible implementations, computer equipment can also determine VaR (Value at Risk). VaR, literally meaning "value at risk," refers to the maximum possible loss of a financial asset or portfolio under normal market fluctuations. More precisely, it refers to the maximum possible loss of the value of a financial asset or portfolio within a specific future period at a certain probability level (pre-set confidence level). It can be defined by the formula: prob[ΔV≥VaR(T, X%)]=1-X%; where prob represents the probability that the asset value loss is less than the maximum possible loss; ΔV is the value loss of a financial asset or portfolio within a certain holding period T; VaR is the value at risk at a confidence level of X% at time T, i.e., the maximum possible loss; and X% is the pre-set confidence level, referring to the given probability, i.e., the confidence level. Commonly used confidence levels in calculating VaR are 95%, 97.5%, or 99%. If we choose 95%, such as... Figure 5 As shown (the horizontal axis represents the range of portfolio value changes, and the vertical axis represents the probability of such changes occurring), this diagram illustrates VaR provided in one embodiment of this application. That is, to... Figure 5 Find the position indicated by the downward arrow on curve 51. At this position, 95% of the value change falls to the right and 5% to the left. The value on the horizontal axis at this position is the VaR value. VaR is typically calculated using years as the unit of time for determining the probability of loss.

[0258] Because it does not meet the subadditivity and consistency conditions of risk measurement, VaR is not a reasonable risk measurement model, and therefore has certain problems in measuring economic capital. Tail VaR fully utilizes the widespread application of VaR in the industry and has well-established hardware and software measures, while also fully considering the possibility of different distributions of loss rates in reality. Therefore, using the Tail VaR method to construct an economic capital measurement model is not only easier to understand, but also more in line with reality, and is theoretically more reasonable than VaR.

[0259] One possible implementation is to simultaneously utilize VaR and TVaR to analyze the predicted balance of the health insurance fund, and combine this with the decision-makers' preferences or utility functions to arrive at the final selection strategy. This process can be repeated until the decision-makers are convinced of the superiority of a particular strategy.

[0260] This application's embodiments are based on dynamic holistic factor analysis for predicting the risk of medical insurance funds. By conducting multiple dynamic simulations based on the increase and consumption of medical insurance funds, combined with the estimated number of outpatient visits and medical expenses, the risk factors and macro-operation of medical insurance funds can be analyzed dynamically and stochastically. At the same time, the introduction of Copula theory can simulate public health emergencies, and by using risk value and tail risk value analysis methods, combined with the preferences or utility functions of decision-makers, scientific and objective macro-decision results can be obtained.

[0261] In the technical solution provided in this application, a target connection function is introduced to assign values ​​to random parameters. Since the target connection function is used to indicate the correlation between random parameters and / or the correlation between random parameters and input parameters, the random parameters and input parameters are parameters related to increase prediction or consumption prediction. When a sudden event occurs, it will inevitably affect increase and consumption. This application considers the correlation between increase and consumption in advance during prediction, which is equivalent to considering the impact of sudden events during prediction, thereby making the final increase and consumption prediction results more accurate.

[0262] By estimating the number of outpatient visits and medical expenses, the increase and depletion of the medical insurance fund can be dynamically predicted, rather than simply determining the future number of outpatient visits and future medical expenses based on historical outpatient visits and medical expenses. The embodiments of this application can make the final predicted increase and depletion of the medical insurance fund more reasonable.

[0263] It should be noted that the examples of input parameters and random parameters in the above embodiments are merely illustrative. In other possible implementations, there may be other types of input parameters and random parameters, which are not limited in this application.

[0264] Please refer to Figure 6 This illustrates a flowchart of a resource allocation method provided in another embodiment of this application. The method can be executed interactively by a terminal and a server, and may include the following steps:

[0265] Step 601: The terminal obtains the input parameters related to the increase and consumption of the medical insurance fund entered in the prediction interface.

[0266] like Figure 7As shown, the user enters input parameters in the prediction interface 70, which displays input boxes for the parameters "Total number of insured urban employees", "Employment ratio", and "Average wage". After the terminal confirms that the user has entered all the above input parameters, the terminal sends a prediction request to the server.

[0267] Step 602: The terminal sends a prediction request to the server. The prediction request is used to request a prediction of the increase and consumption of the medical insurance fund, and the prediction request carries input parameters.

[0268] Step 603: The server determines the values ​​of k sets of random parameters for the prediction model based on the target relevance.

[0269] Where k is a positive integer, the random parameter refers to the randomly generated parameter related to the prediction of the increase and consumption of the medical insurance fund; the target correlation includes at least one of the following: the correlation between random parameters, the correlation between random parameters and input parameters.

[0270] Step 604: The server processes the values ​​of the input parameters and random parameters through the prediction model to obtain k initial prediction results of the increase in medical insurance funds and k initial prediction results of the consumption of medical insurance funds.

[0271] Step 605: Based on the initial prediction results of the increase in medical insurance funds and the initial prediction results of the consumption of medical insurance funds in k groups, the server determines the final prediction results of the increase in medical insurance funds and the final prediction results of the consumption of medical insurance funds.

[0272] Step 606: The server sends the final predicted increase in medical insurance funds and the final predicted consumption of medical insurance funds to the terminal.

[0273] Accordingly, the terminal receives the final predicted increase in medical insurance funds and the final predicted consumption of medical insurance funds from the server.

[0274] Step 607: The terminal displays the final predicted increase in medical insurance funds and the final predicted consumption of medical insurance funds on the prediction interface.

[0275] like Figure 7 As shown, the prediction interface 70 displays the final prediction result 71 of the increase in medical insurance funds and the final prediction result 72 of the consumption of medical insurance funds.

[0276] In possible implementations, the predicted increase in medical insurance funds can include predictions of the increase in medical insurance funds over one year or more.

[0277] In some possible implementations, the medical insurance fund consumption forecast results may include medical insurance fund consumption forecast results for one year or more.

[0278] In one possible implementation, the user can also modify the value of the input parameter in the prediction interface. The terminal obtains a reset command corresponding to the input parameter; based on the reset command, the terminal obtains the updated input parameter entered in the prediction interface.

[0279] like Figure 7 As shown, the prediction interface 70 displays a reset control 73, which the user can touch to trigger a reset command to be sent to the terminal.

[0280] The following are embodiments of the apparatus described in this application, which can be used to execute the embodiments of the method described in this application. For details not disclosed in the apparatus embodiments of this application, please refer to the embodiments of the method described in this application.

[0281] Please refer to Figure 8 This diagram illustrates a block diagram of a resource allocation apparatus according to an embodiment of this application. The apparatus has the functionality to implement the method example described above; this functionality can be implemented in hardware or by hardware executing corresponding software. The apparatus can be a computer device or can be installed within a computer device. The apparatus 800 may include: a parameter acquisition module 810, a parameter generation module 820, a parameter processing module 830, and an information determination module 840.

[0282] The parameter acquisition module 810 is used to acquire input parameters related to the prediction of the number of outpatient visits of the medical institution. The input parameters refer to the user-input parameters related to the prediction of the number of outpatient visits of the medical institution.

[0283] The parameter generation module 820 is used to determine k sets of random parameters based on the input parameters and the target correlation. The random parameters refer to randomly generated parameters related to the predicted number of outpatient visits of the medical institution. The target correlation includes the correlation between the random parameters and the input parameters. The k is a positive integer.

[0284] The parameter processing module 830 is used to perform k rounds of visitor prediction processing based on the input parameters and the k sets of random parameters to obtain the k sets of initial visitor prediction results for the medical institution.

[0285] The parameter processing module 830 is also used to determine the final predicted number of visits to the medical institution based on the initial predicted number of visits for the k groups.

[0286] The information determination module 840 is used to determine resource allocation suggestion information for the medical institution based on the final predicted number of outpatient visits. The resource allocation suggestion information is used to provide suggestions when allocating resources to the medical institution.

[0287] In summary, the technical solution provided in this application determines the values ​​of multiple sets of random parameters based on the correlation between input parameters and random parameters. Then, based on the input parameters and multiple sets of random parameters, multiple initial prediction results of patient visits are determined. Finally, based on the multiple initial prediction results of patient visits, the final prediction result of patient visits is determined. This application combines consideration of the correlation between various parameters and assigns values ​​to random parameters based on this correlation. By randomly simulating the distribution of possible result ranges (i.e., the aforementioned multiple sets of initial prediction results of patient visits), rather than just estimating results at a single point, the volatility of medium- and long-term predictions is considered, improving the accuracy of the final prediction result of patient visits. Based on the final prediction result of patient visits for the medical institution, resource allocation recommendations for the medical institution are determined. Since patient visits are fully considered during resource allocation, the resource allocation of the medical institution is more reasonable, maximizing resource utilization efficiency.

[0288] In an illustrative embodiment, the parameter generation module 820 is used for:

[0289] Based on historical input parameters and historical random parameters, determine the marginal distribution functions of the historical input parameters and the historical random parameters, respectively.

[0290] Based on the marginal distribution function of the historical input parameters and the historical random parameters, a target connection function is obtained, which is used to indicate the target relevance;

[0291] Based on the target connection function, the k sets of random parameters are determined.

[0292] In an illustrative embodiment, the information determination module 840 is used for:

[0293] Based on the final predicted number of outpatient visits to the medical institution, the predicted resource demand information of the medical institution is determined. The predicted resource demand information refers to the predicted resource demand of the medical institution.

[0294] Based on the resource demand forecast information of the medical institution, resource allocation recommendations are determined for the medical institution.

[0295] In an illustrative embodiment, the resource demand forecasting information includes equipment demand forecasting information and / or physician demand forecasting information. The equipment demand forecasting information refers to the predicted equipment demand of the medical institution, and the physician demand forecasting information refers to the predicted physician demand of the medical institution.

[0296] The information determination module 840 is used for:

[0297] The final predicted number of outpatient visits is determined by identifying the target number of outpatient visits within the i-th age range and at the j-th type of medical institution, where i and j are positive integers.

[0298] Based on the predicted target number of outpatient visits and the target correspondence, the predicted resource demand information is determined. The target correspondence refers to the correspondence between users in the i-th age range who visit the j-th type of medical institution and their equipment demand and / or the correspondence between users in the i-th age range who visit the j-th type of medical institution and their physician demand.

[0299] In an illustrative embodiment, the parameter processing module 830 is used for:

[0300] The input parameters and the k sets of random parameters are processed by a machine learning model to obtain the k sets of initial predicted visitor numbers. The machine learning model refers to a neural network model used to predict the visitor numbers of the medical institution.

[0301] In an illustrative embodiment, the input parameters include the total number of insured persons; the random parameters include the number of annual outpatient visits by individuals within the i-th age range and who visit medical institutions of the j-th type, where i and j are positive integers;

[0302] The parameter processing module 830 is used for:

[0303] The total number of insured persons is divided into n categories of insured persons according to n age ranges using a prediction model, where n is a positive integer greater than or equal to i;

[0304] Based on the individual's annual number of outpatient visits and the number of people in each of the n categories of insured persons, the initial predicted number of outpatient visits for the k groups is determined.

[0305] In an illustrative embodiment, the parameter processing module 830 is further configured to:

[0306] Based on the initial predicted number of outpatient visits for the k groups, the initial predicted increase in medical insurance funds and the initial predicted consumption of medical insurance funds for the k groups are determined.

[0307] Based on the initial predictions of the increase in medical insurance funds and the initial predictions of the consumption of medical insurance funds in the k groups, the final predictions of the increase in medical insurance funds and the final predictions of the consumption of medical insurance funds are determined.

[0308] In an illustrative embodiment, the device further includes a value determination module (not shown in the figure).

[0309] Based on the final predicted increase in the medical insurance fund and the final predicted depletion of the medical insurance fund, the final predicted balance of the medical insurance fund is determined.

[0310] The risk value of the final predicted balance of the medical insurance fund is determined at a pre-set confidence level, and the risk value is used as a threshold to indicate the degree of loss.

[0311] In an illustrative embodiment, the value determination module is used for:

[0312] Based on the preset confidence level, an initial value at risk is determined for the loss amount of the final predicted medical insurance fund balance. The initial value at risk is used to indicate the maximum possible loss of the final predicted medical insurance fund balance.

[0313] Obtain the distribution density of the loss amount in the final predicted result of the medical insurance fund balance;

[0314] The risk value is determined based on the distribution density of the loss amount from the final predicted medical insurance fund balance, the initial value at risk, and the pre-set confidence level.

[0315] It should be noted that the apparatus provided in the above embodiments is only illustrated by the division of the above functional modules when implementing its functions. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the content structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the apparatus and method embodiments provided in the above embodiments belong to the same concept, and the specific implementation process can be found in the method embodiments, which will not be repeated here.

[0316] Please refer to Figure 9 This illustration shows a schematic diagram of a computer device 900 provided in one embodiment of this application. The computer device 900 can be used to implement the resource allocation method on the computer device side provided in the above embodiments. The computer device 900 can be... Figure 1 The terminal 10 or server 20 described in the embodiment. Specifically:

[0317] The computer device 900 includes a central processing unit (CPU) 901, a system memory 904 including RAM (Random Access Memory) 902 and ROM (Read-Only Memory) 903, and a system bus 905 connecting the system memory 904 and the CPU 901. The computer device 900 also includes a basic input / output system (I / O system) 906 that facilitates information transfer between various components within the computer, and a mass storage device 907 for storing the operating system 913, application programs 914, and other program modules 915.

[0318] The basic input / output system 906 includes a display 908 for displaying information and an input device 909 for user input, such as a mouse or keyboard. Both the display 908 and the input device 909 are connected to the central processing unit 901 via an input / output controller 910 connected to the system bus 905. The basic input / output system 906 may also include the input / output controller 910 for receiving and processing input from multiple other devices such as a keyboard, mouse, or electronic stylus. Similarly, the input / output controller 910 also provides output to a display screen, printer, or other types of output devices.

[0319] The mass storage device 907 is connected to the central processing unit 901 via a mass storage controller (not shown) connected to the system bus 905. The mass storage device 907 and its associated computer-readable media provide non-volatile storage for the computer device 900. That is, the mass storage device 907 may include computer-readable media (not shown) such as a hard disk or a CD-ROM (Compact Disc Read-Only Memory) drive.

[0320] Without loss of generality, the computer-readable medium may include computer storage media and communication media. Computer storage media include volatile and non-volatile, removable and non-removable media implemented using any method or technology for storing information such as computer-readable instructions, data structures, program modules, or other data. Computer storage media include RAM, ROM, EPROM (Erasable Programmable Read-Only Memory), EEPROM (Electrically Erasable Programmable Read-Only Memory), flash memory or other solid-state storage devices, CD-ROM, DVD (Digital Versatile Disc) or other optical storage, magnetic tape cassettes, magnetic tape, disk storage, or other magnetic storage devices. Of course, those skilled in the art will recognize that the computer storage media are not limited to the above-mentioned types. The system memory 904 and mass storage device 907 described above can be collectively referred to as memory.

[0321] According to various embodiments of this application, the computer device 900 can also be connected to a remote computer on a network, such as the Internet. That is, the computer device 900 can be connected to a network 912 via a network interface unit 911 connected to the system bus 905, or the network interface unit 911 can be used to connect to other types of networks or remote computer systems (not shown).

[0322] The memory also includes one or more programs stored in the memory and configured to be executed by one or more processors. These programs contain instructions for implementing the resource allocation method on the computer device side.

[0323] In an illustrative embodiment, a computer device is also provided, comprising a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set, or an instruction set. The at least one instruction, at least one program, code set, or instruction set is configured to be executed by one or more processors to implement the resource allocation method described above on the computer device side.

[0324] In an illustrative embodiment, a computer-readable storage medium is also provided, which stores at least one instruction, at least one program, a code set, or an instruction set, wherein the at least one instruction, the at least one program, the code set, or the instruction set implements the above-described resource allocation method when executed by a processor of a computer device.

[0325] In an exemplary embodiment, a computer program product or computer program is also provided, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the resource allocation method described above.

[0326] It should be understood that "multiple" as used herein refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. Furthermore, the step numbers described herein are merely illustrative of one possible execution order. In some other embodiments, the steps may not be executed in numerical order, such as two steps with different numbers being executed simultaneously, or two steps with different numbers being executed in the reverse order of the illustration. This application does not limit this.

[0327] Those skilled in the art will understand that all or part of the steps of the above embodiments can be implemented by hardware or by a program instructing related hardware. The program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk.

[0328] The above description is merely an exemplary embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A resource allocation method, characterized in that, The method includes: Obtain input parameters related to the prediction of the number of outpatient visits to the medical institution, wherein the input parameters refer to user-input parameters related to the prediction of the number of outpatient visits to the medical institution; Based on the input parameters and target correlation, k sets of random parameters are determined. The random parameters refer to randomly generated parameters that are related to the predicted number of outpatient visits to the medical institution. The target correlation includes the correlation between the random parameters and the input parameters. The k is a positive integer. Based on the input parameters and the k sets of random parameters, k rounds of outpatient visit prediction processing are performed to obtain the initial outpatient visit prediction results for the medical institution. Based on the initial patient visit prediction results of the k groups, the final patient visit prediction results of the medical institution are determined. The final predicted number of outpatient visits is determined by identifying the target number of outpatient visits within the i-th age range and at the j-th type of medical institution, where i and j are positive integers. Based on the predicted number of outpatient visits and the target correspondence, resource demand prediction information is determined; wherein, the target correspondence refers to the correspondence between users in the i-th age range who visit the j-th type of medical institution and their equipment demand and / or the correspondence between users in the i-th age range who visit the j-th type of medical institution and their physician demand; the resource demand prediction information includes equipment demand prediction information and / or physician demand prediction information, wherein the equipment demand prediction information refers to the predicted equipment demand of the medical institution, and the physician demand prediction information refers to the predicted physician demand of the medical institution. Based on the resource demand forecast information of the medical institution, resource allocation recommendation information is determined for the medical institution. The resource allocation recommendation information is used to provide suggestions when allocating resources to the medical institution.

2. The method according to claim 1, characterized in that, The step of determining k sets of random parameters based on the input parameters and target relevance includes: Based on historical input parameters and historical random parameters, determine the marginal distribution functions of the historical input parameters and the historical random parameters, respectively. Based on the marginal distribution function of the historical input parameters and the historical random parameters, a target connection function is obtained, which is used to indicate the target relevance; Based on the target connection function, the k sets of random parameters are determined.

3. The method according to claim 1, characterized in that, The process of predicting patient visits for k rounds based on the input parameters and the k sets of random parameters yields k initial prediction results for patient visits at the medical institution, including: The input parameters and the k sets of random parameters are processed by a machine learning model to obtain the k sets of initial predicted visitor numbers. The machine learning model refers to a neural network model used to predict the visitor numbers of the medical institution.

4. The method according to claim 1, characterized in that, The input parameters include the total number of insured persons; the random parameters include the number of annual outpatient visits by individuals who are in the i-th age range and seek medical treatment at the j-th type of medical institution. The process of predicting patient visits for k rounds based on the input parameters and the k sets of random parameters yields k initial prediction results for patient visits at the medical institution, including: The total number of insured persons is divided into n categories of insured persons according to n age ranges using a prediction model, where n is a positive integer greater than or equal to i; Based on the individual's annual number of outpatient visits and the number of people in each of the n categories of insured persons, the initial predicted number of outpatient visits for the k groups is determined.

5. The method according to claim 1, characterized in that, After performing k rounds of visitor prediction processing based on the input parameters and the k sets of random parameters to obtain the initial visitor prediction results for the medical institution, the process further includes: Based on the initial predicted number of outpatient visits for the k groups, the initial predicted increase in medical insurance funds and the initial predicted consumption of medical insurance funds for the k groups are determined. Based on the initial predictions of the increase in medical insurance funds and the initial predictions of the consumption of medical insurance funds in the k groups, the final predictions of the increase in medical insurance funds and the final predictions of the consumption of medical insurance funds are determined.

6. The method according to claim 5, characterized in that, The process of determining the final predicted increase and consumption of the medical insurance fund based on the k initial predicted increases and consumption results of the medical insurance fund also includes: Based on the final predicted increase in the medical insurance fund and the final predicted depletion of the medical insurance fund, the final predicted balance of the medical insurance fund is determined. The risk value of the final predicted balance of the medical insurance fund is determined at a pre-set confidence level, and the risk value is used as a threshold to indicate the degree of loss.

7. The method according to claim 6, characterized in that, Determining the risk value of the final predicted medical insurance fund balance at a pre-set confidence level includes: Based on the preset confidence level, an initial value at risk is determined for the loss amount of the final predicted medical insurance fund balance. The initial value at risk is used to indicate the maximum possible loss of the final predicted medical insurance fund balance. Obtain the distribution density of the loss amount in the final predicted result of the medical insurance fund balance; The risk value is determined based on the distribution density of the loss amount from the final predicted medical insurance fund balance, the initial value at risk, and the pre-set confidence level.

8. A resource allocation device, characterized in that, The device includes: The parameter acquisition module is used to acquire input parameters related to the prediction of the number of outpatient visits to the medical institution. The input parameters refer to the user-input parameters related to the prediction of the number of outpatient visits to the medical institution. The parameter generation module is used to determine k sets of random parameters based on the input parameters and the target correlation. The random parameters refer to randomly generated parameters related to the predicted number of outpatient visits of the medical institution. The target correlation includes the correlation between the random parameters and the input parameters. The k is a positive integer. The parameter processing module is used to perform k rounds of visitor prediction processing based on the input parameters and the k sets of random parameters to obtain the k sets of initial visitor prediction results for the medical institution. The parameter processing module is also used to determine the final predicted number of visits to the medical institution based on the initial predicted number of visits for the k groups. The information determination module is used to determine the target number of visits predicted in the final visit prediction results, which are within the i-th age range and visit the j-th type of medical institution, where i and j are positive integers; The information determination module is further configured to determine resource demand prediction information based on the target number of outpatient visits prediction results and the target correspondence; wherein, the target correspondence refers to the correspondence between users in the i-th age range who are visiting the j-th type of medical institution and their equipment demand and / or the correspondence between users in the i-th age range who are visiting the j-th type of medical institution and their physician demand; the resource demand prediction information includes equipment demand prediction information and / or physician demand prediction information, wherein the equipment demand prediction information refers to the predicted equipment demand of the medical institution, and the physician demand prediction information refers to the predicted physician demand of the medical institution; The information determination module is further configured to determine resource allocation recommendation information for the medical institution based on the resource demand prediction information of the medical institution, and the resource allocation recommendation information is used to provide suggestions when allocating resources to the medical institution.

9. A computer device, characterized in that, The computer device includes a processor and a memory, the memory storing at least one instruction, at least one program, a code set, or an instruction set, the at least one instruction, the at least one program, the code set, or the instruction set being loaded and executed by the processor to implement the resource allocation method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The storage medium stores at least one instruction, at least one program, code set, or instruction set, wherein the at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by a processor to implement the resource allocation method as described in any one of claims 1 to 7.

11. A computer program product, characterized in that, The computer program product includes computer instructions that are executed by a processor to implement the resource allocation method as described in any one of claims 1 to 7.

Images