A Credit-Based Medical Limit Evaluation Method Based on Decision Enhancement Tree

Abstract

This invention relates to the field of medical technology and discloses a credit-based medical limit evaluation method based on a decision enhancement tree. By adjusting the initial limit according to the patient's illness, it can more accurately match the patient's medical needs and avoid a "one-size-fits-all" limit setting. For patients with high rehabilitation costs, the credit-based medical limit evaluation system can provide some financial support and alleviate their financial burden. By scoring patients' overdue payments during hospitalization and overdue payments outside of hospitalization differently, this method can more intuitively reflect the patient's credit status, making the impact of overdue payments during hospitalization on the patient's credit limit smaller. This method can convey a positive incentive signal, but it also understands and tolerates overdue behavior in special circumstances. This inclusiveness helps to enhance patients' credit awareness and encourages them to more actively maintain their credit records in the future.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of credit medical treatment, in particular to a credit medical treatment quota evaluation method based on a decision boosting tree. BACKGROUND

[0002] Credit medical treatment is a medical treatment mode in which a certain quota is granted to a patient after the credit of the patient is evaluated, and the patient realizes the payment after diagnosis and treatment, and the medical expenses generated during the treatment can be deducted using the credit quota first, and the patient completes the payment once within a certain period of time after the treatment.

[0003] Adopting credit medical treatment can mainly avoid the situation that the patient has to go through the payment link before the subsequent diagnosis and treatment process in one treatment process, reduce the payment link and queuing, and reduce the short-term financial payment pressure of the patient.

[0004] The existing credit medical treatment mode usually uses a single reference index to evaluate the credit of the patient, and only uses the social history and economic situation of the patient to evaluate the quota of the patient, and this evaluation method cannot give the patient a quota according to the actual use demand of the patient during the medical treatment, which will lead to the situation that the supply and demand are uneven. Especially for some patients who need long-term rehabilitation, when the granted quota is used up, they still need to queue and pay by themselves, which will lead to the fact that the patient cannot enjoy the convenience brought by credit medical treatment. Therefore, a credit medical treatment quota evaluation method based on a decision boosting tree is proposed to solve the above problems. SUMMARY

[0005] Technical problems to be solved

[0006] In view of the defects in the prior art, the present application provides a credit medical treatment quota evaluation method based on a decision boosting tree, which has the advantages of reducing the payment link and queuing of the patient during the treatment, reducing the short-term financial payment pressure of the patient, and solving the problem that the patient has to go through the payment link before the subsequent diagnosis and treatment process in one treatment process.

[0007] Technical scheme

[0008] To achieve the above purpose, the present application provides the following technical scheme: a credit medical treatment quota evaluation method based on a decision boosting tree, comprising the following steps,

[0009] S1, a database DB for storing patient information is established, the patient registers for outpatient service and submits personal information to the database DB, the existing patient data in the database DB and the submitted personal information of the patient are matched, it is judged whether the patient has a credit quota, if yes, S3 is entered, if not, S2 is entered;

[0010] S2, a set of social history evaluation factors of the patient is obtained to construct an initial credit evaluation system to grant the patient an initial credit limit and enter S3;

[0011] S3, the patient makes an appointment, the doctor diagnoses the patient's condition, and divides it into short-term rehabilitation cases and long-term rehabilitation cases according to the rehabilitation time of the condition If the patient is a short-term rehabilitation case , enter S5, if the patient is a long-term rehabilitation case , enter S4;

[0012] S4, according to the long-term rehabilitation cases of the patient , construct a one-time credit adjustment evaluation system, and adjust the credit limit of the patient to a credit limit through the one-time credit adjustment evaluation system ;

[0013] The long-term rehabilitation cases in S4 include different disease types m that need long-term rehabilitation, each disease type m that needs long-term rehabilitation is divided into different severity levels y, different severity levels y correspond to different rehabilitation times k, each different disease type m corresponds to a different initial credit limit , each different disease type m corresponds to a different initial weight , define a mapping function :

[0014] ;

[0015] In order to avoid the value being too large or too small, the output value of the mapping function is intercepted to make its output value between the interval , through the formula:

[0016] ;

[0017] Where is the minimum output value of the mapping function, is the maximum output value of the mapping function, the weight is adjusted according to the mapping function , to obtain the adjusted weight :

[0018] ;

[0019] According to the adjusted weight and the initial credit limit of each m , the adjustment limit is obtained : ​

[0020] ;

[0021] To determine the adjustment limit And the initial credit limit Adjust the credit limit to obtain a secondary credit limit. :

[0022] ;

[0023] in , , For coefficients, For patients with disease m, the shortest recovery time. The longest recovery time for a patient with disease m;

[0024] S5, patients receive treatment based on the doctor's diagnosis, and the resulting treatment costs are covered by the credit limit. Making advance payments generates a used credit limit. And obtain the patient's treatment history;

[0025] S6. Settlement will be conducted after treatment is completed, and the settlement will be based on the patient's medical insurance reimbursement limit. To achieve quota settlement and obtain the settlement quota. ;

[0026] S7. Based on the patient's medical records and the actual settlement amount. To build a machine learning model to update the credit limit adjustment system;

[0027] S8. Patient's settlement amount Repayment will be made based on the patient's history of good repayment record. Establish a secondary credit limit adjustment evaluation system to determine the credit limit. The system stores the patient's credit limit information in the database DB and updates the existing patient information.

[0028] Furthermore, the social resume evaluation factor set in S2 It includes x evaluation factors, each evaluation factor x has h rating levels, each evaluation factor x is assigned a weight w, and each rating level h is assigned a score j; through the formula:

[0029] ;

[0030] To obtain the patient's social history score F, the range of values ​​for the social history score F is determined. Divide into four sub-intervals from smallest to largest:

[0031] ;

[0032] in Each corresponds to ,in Each corresponds to a different credit limit, and According to the formula:

[0033] ;

[0034] according to To determine the credit limit .

[0035] Furthermore, the settlement amount in S6 According to the formula:

[0036] ;

[0037] in For settlement amount, This refers to the reimbursement limit under medical insurance; when a patient does not use medical insurance, =0.

[0038] further, For a patient with disease type m and severity level y, the actual settlement amount required is estimated by establishing a gradient boosting tree model to predict the patient's treatment costs. And by comparing the estimated costs With credit limit Difference value , to Adjust the value range and construct the initial model:

[0039] ;

[0040] in It is the sample size. This is the actual treatment cost for the i-th sample. This is the initial estimated value;

[0041] Calculate the residuals: For each sample i, calculate the residuals between the current model prediction and the actual value:

[0042] ;

[0043] in, The actual treatment cost for sample i. The estimated cost for sample i, where It is the feature vector of the i-th sample;

[0044] Use decision trees to fit the residuals. To obtain the predicted value of the t-th tree. ;

[0045] ;

[0046] Where L is the loss function. For trees Learning rate The number of decision trees;

[0047] The new decision tree is added to the existing prediction model to update the overall prediction and obtain the final prediction model:

[0048] ;

[0049] To predict the treatment cost of a new sample, based on: To derive the predicted treatment cost With credit limit deviation value ,according to Update the mapping function by its size The range of values ​​for , where when >0, Decrease, when When <0, Increase, according to:

[0050] ;

[0051] Come to Adjustments are made, among which It is based on The adjustment step size is determined by the distribution and size of the elements. This is an adjustment function.

[0052] Furthermore, the tree is analyzed based on the number of disease types m that the patients suffer from. learning rate Adjustments are made using the following formula:

[0053] ;

[0054] Where k is the total number of different diseases the patient has, and i is the i-th disease the patient has, i = (1, 2, ..., k). It represents the level corresponding to the i-th disease of the patient. and These are hyperparameters used to control the impact of disease severity and the number of diseases on the learning rate. It is an offset of the base value used to adjust the learning rate.

[0055] Furthermore, the patient's credit history for subsequent repayments in S8. Including the number of overdue payments Next time, overdue period Days, and on-time repayment rate Each evaluation factor is assigned a score, and the total score is calculated, based on:

[0056] ;

[0057] Credit limit Adjust to credit limit ,in For each overdue point, For points awarded if the deadline is one day overdue, The initial score; 10 For every 10% increase in on-time repayment rate, point, Where j is the weight, representing the weight during the patient's hospitalization. , Weighting for patients outside of hospitalization.

[0058] Furthermore, the aforementioned repayment on-time rate According to the formula:

[0059] ;

[0060] in For loans that have matured and been repaid, These are loans that have already matured.

[0061] Beneficial effects

[0062] Compared with existing technologies, this invention provides a credit medical limit evaluation method based on decision boosting tree, which has the following beneficial effects:

[0063] 1. This credit-based medical credit limit evaluation method, based on a decision enhancement tree, adjusts credit limits according to the patient's medical records, enabling a more precise match to the patient's medical needs and avoiding a "one-size-fits-all" approach to credit limit settings. For patients with high rehabilitation costs, the credit-based medical credit limit evaluation system can provide some financial support, alleviating their financial burden.

[0064] 2. This credit-based medical limit evaluation method, based on a decision enhancement tree, scores patients' overdue payments during hospitalization and those outside of hospitalization differently. This approach more directly reflects a patient's creditworthiness, minimizing the impact of overdue payments during hospitalization on their credit limit. This method sends a positive incentive signal, encouraging patients to be honest and trustworthy while understanding and tolerating overdue behavior in special circumstances. This inclusiveness helps enhance patients' credit awareness and encourages them to more actively maintain their credit records in the future. Attached Figure Description

[0065] Figure 1 This is a flowchart of a credit medical limit evaluation method based on decision boosting tree proposed in this invention;

[0066] Figure 2 This is a schematic diagram of social resume evaluation factors for a credit medical limit evaluation method based on decision enhancement tree proposed in this invention.

[0067] Figure 3 This is a schematic diagram of a long-term rehabilitation case for a credit medical limit evaluation method based on decision boosting tree proposed in this invention. Detailed Implementation

[0068] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0069] Please see Figure 1 To avoid patients having to go through a payment process before proceeding with subsequent treatments during a single visit, this initiative aims to reduce payment steps and waiting times, thereby alleviating short-term financial burdens on patients.

[0070] This embodiment provides a credit medical limit evaluation method based on decision enhancement tree, including the following steps:

[0071] S1. Establish a database DB for storing patient information. Patients register for outpatient services and submit their personal information to the database DB. The database DB is matched with the submitted patient information to determine whether the patient has a credit limit. If yes, proceed to S3; otherwise, proceed to S2.

[0072] S2, based on the patient's social history evaluation factor set To construct an initial credit limit evaluation system to grant patients an initial credit limit. And enter S3;

[0073] S3: The patient seeks medical attention, the doctor diagnoses the patient's condition, and classifies the patient as a short-term recovery case based on the recovery time. and long-term rehabilitation cases If the patient is a short-term recovery case Entering S5, if the patient is a long-term recovery case Then proceed to S4;

[0074] S4, based on the patient's long-term recovery case To construct a credit limit adjustment evaluation system, and to adjust the patient's credit limit through this system. Adjust to credit limit ;

[0075] S5, patients receive treatment based on the doctor's diagnosis, and the resulting treatment costs are covered by the credit limit. Making advance payments generates a used credit limit. ;

[0076] S6. Settlement will be conducted after treatment is completed, and the settlement will be based on the patient's medical insurance reimbursement limit. To achieve quota settlement and obtain the settlement quota. ;

[0077] S7. Based on the patient's medical records and the actual settlement amount. To build a machine learning model to update the credit limit adjustment system;

[0078] S8. Patient's settlement amount Repayment will be made based on the patient's history of good repayment record. Establish a secondary credit limit adjustment evaluation system to determine the credit limit. The patient's credit limit information is stored in the database DB, and the original data is updated.

[0079] During the actual medical visit, patients can simultaneously provide personal information to the database (DB) for matching with existing user information, such as their ID card number, to ensure accurate data matching. If the database does not contain corresponding credit limit data for the patient, the user can choose whether to participate in credit-based medical care. When choosing to participate, the patient can submit their personal social history to obtain a credit limit. Once the credit limit is obtained, subsequent consultations will be conducted. Doctors will diagnose the patient's condition. For cases with short-term recoverable illnesses, such as colds, fevers, or diarrhea, patients can directly seek medical attention, and all treatment costs incurred during the consultation will be covered by the credit limit. However, for cases requiring long-term recovery... In this case, the credit limit of the patient is adjusted according to the type of disease m, the severity level of the disease y, and the recovery time j, so as to meet the needs of patients who need long-term treatment and avoid their inability to enjoy the convenience of credit medical care during the treatment process due to insufficient credit limit.

[0080] The above methods can help patients avoid having to pay before proceeding with subsequent treatments during a single visit, reducing payment steps and waiting times, and alleviating short-term financial burdens on patients.

[0081] At the same time, the effect of adjusting the patient's credit limit three times can ensure that patients with chronic diseases are given the appropriate credit limit, avoiding the inability to enjoy the convenience of credit medical care due to insufficient credit limit.

[0082] according to Figure 2 Social resume evaluation factor set in S1 It includes x evaluation factors, each evaluation factor x has h rating levels, each evaluation factor x is assigned a weight w, and each rating level h is assigned a score j using the formula:

[0083] ;

[0084] To obtain the patient's social history score F, the range of values ​​for the social history score F is determined. Divide it into four sub-intervals from smallest to largest:

[0085] ;

[0086] in Each corresponds to ,in Each corresponds to a different credit limit, and According to the formula:

[0087] ;

[0088] according to To determine the credit limit .

[0089] The evaluation factors in this embodiment include the patient's age, income, occupation, and education level. By considering the patient's individual characteristics, such as age, income, and occupation, their economic capacity and needs can be more accurately assessed, thereby providing more personalized service or cost support plans. This helps ensure that patients can receive rehabilitation services that match their economic situation and alleviate their financial burden.

[0090] As a preferred technical solution in this embodiment, in actual use, it can be... The initial amounts are set at 3000 yuan, 5000 yuan, 8000 yuan, and 10000 yuan respectively. Setting these initial amounts can meet the needs of patients for routine medical visits. These amounts represent only a reasonable scenario in practice and do not represent all initial amounts. The initial amounts can be set according to actual usage, depending on hospital charges and regional differences.

[0091] Please see Figure 3 The long-term rehabilitation cases in S4 of this embodiment This includes different types of diseases m that require long-term rehabilitation. Each type of disease m requiring long-term rehabilitation is divided into different severity levels y. Different severity levels y correspond to different rehabilitation times k. Each different type of disease m has a different initial allowance. Each different disease m corresponds to a different initial weight. Based on the above adjustment factors, define the mapping function. :

[0092] ;

[0093] In order to avoid If the value is too large or too small, it affects the mapping function. The output value is truncated to make it fall within the range Between them, through the formula:

[0094] ;

[0095] in The minimum output value of the mapping function. The maximum output value of the mapping function, based on the mapping function. To adjust the weight The adjusted weights are obtained. :

[0096] ;

[0097] Based on the adjusted weights and the initial credit limit for each m To determine the adjustment limit :

[0098] ;

[0099] To determine the adjustment limit And for a single credit limit Adjust the credit limit to obtain a secondary credit limit. :

[0100] ;

[0101] in , , For coefficients, For patients with disease m, the shortest recovery time. The longest recovery time for a patient with disease m. This includes cases of short-term recovery. =0.

[0102] Furthermore, the settlement amount in S6 According to the formula:

[0103] ;

[0104] in For settlement amount, This refers to the reimbursement limit under medical insurance; when a patient does not use medical insurance, =0.

[0105] In the actual evaluation process, the selected m may include cases requiring long-term rehabilitation, such as those with neurological injuries, bone and joint injuries, cardiovascular diseases, and chronic respiratory diseases. Of course, the above cases are only a portion of long-term rehabilitation cases, and the number of cases or severity levels can be increased according to actual needs. The severity level y can be divided into four levels: Level 1 (mild), Level 2 (moderate), Level 3 (severe), and Level 4 (extremely severe). Similarly, the severity level can be further refined in the actual evaluation process to ensure accuracy. Furthermore, different rehabilitation times k can be assigned based on the severity level y. , , The coefficient is determined through expert evaluation based on factors such as the patient's physical condition, the hospital's medical level, and geographical location. Of course, the above factors only represent a portion of those that may influence disease type m. In practical application, the coefficient can be determined based on more factors that can affect disease type m. , , Equal coefficients.

[0106] Furthermore, when a patient suffers from two chronic diseases, the credit limit adjustment process should involve two evaluations of the patient through the credit limit adjustment system to determine the final adjusted credit limit. ,in This is the adjustment amount obtained after the second adjustment.

[0107] Given a patient with disease type m and severity level y, a gradient boosting tree model is established to estimate the actual settlement amount required for treatment. And by comparing the estimated costs Difference from credit limit E4 , to Adjust the value range and construct the initial model:

[0108] ;

[0109] in It is the sample size. This is the actual treatment cost for the i-th sample. This is the initial estimated value;

[0110] Calculate the residuals: For each sample i, calculate the residuals between the current model prediction and the actual value:

[0111] ;

[0112] in, The actual treatment cost for sample i. The estimated cost for sample i, where It is the feature vector of the i-th sample;

[0113] Use decision trees to fit the residuals. To obtain the predicted value of the t-th tree. The number of decision trees is 1-t;

[0114] ;

[0115] Where L is the loss function. For trees The learning rate is t, where t is the number of decision trees.

[0116] The new decision tree is added to the existing prediction model to update the overall prediction and obtain the final prediction model:

[0117] ;

[0118] To predict the treatment cost of a new sample, based on: To derive the predicted treatment cost With credit limit deviation value ,according to Update the mapping function by its size The range of values ​​for , where when >0, Decrease, when When <0, Increase, according to:

[0119] ;

[0120] Come to Adjustments are made, among which It is based on The adjustment step size is determined by the distribution and size of the distribution. For adjustment functions. When When >0, The output value is 1 when When =0, The output value is 0 when When <0, The output value is -1.

[0121] in The loss function is used to measure the model's predictions. Compared with the true value The differences between them. It is the sum of predictions made by t-1 trees for the i-th sample. It is the true label of the i-th sample, where This indicates the search for optimal parameters while minimizing the loss function L. The process.

[0122] In the actual evaluation process, the selected m may include cases requiring long-term rehabilitation, such as those with neurological injuries, bone and joint injuries, cardiovascular diseases, and chronic respiratory diseases. Of course, the above cases are only a portion of long-term rehabilitation cases, and the number of cases or severity levels can be increased according to actual needs. The severity level y can be divided into four levels: Level 1 (mild), Level 2 (moderate), Level 3 (severe), and Level 4 (extremely severe). Similarly, the severity level can be further refined in the actual evaluation process to ensure accuracy. Furthermore, different rehabilitation times k can be assigned based on the severity level y. , , The coefficients are determined through expert evaluation based on factors such as the patient's physical condition, the hospital's medical level, and geographical location. Experts possess profound professional knowledge and rich practical experience in their respective fields, resulting in more comprehensive and accurate evaluation results. It is worth noting that... Figure 3 The diseases shown represent only a portion of actual cases. In the actual evaluation process, the diseases can be improved based on the patient's actual condition.

[0123] Preferably, in this embodiment, the learning rate of the tree ht can be determined based on the number of diseases m that the patient suffers from. Adjustments are made using the following formula:

[0124] ;

[0125] Where k is the total number of different diseases the patient suffers from, i is the i-th disease the patient suffers from, i = (1, 2, ..., k), and yi is the level corresponding to the i-th disease of the patient. and These are hyperparameters used to control the impact of disease severity and the number of diseases on the learning rate. It is a base value used to adjust the learning rate, where... It is used as part of the learning rate q formula to reflect the impact of the number of disease types k on credit limit prediction. As the number of disease types increases, the value of log(1+k) also increases, but the rate of increase gradually decreases. This helps to avoid the learning rate q being too high when there are many disease types, thereby maintaining the rationality and stability of the predicted credit limit.

[0126] In the above embodiments, by dynamically adjusting the learning rate q, the scheme can flexibly address the number and severity of different patients' illnesses. This flexibility allows the model to more accurately reflect a patient's credit status, as the number and severity of illnesses are often important factors in determining credit limits. The scheme considers the impact of the number and severity of illnesses on credit limits, making the allocation of credit limits more equitable. Patients with more illnesses or higher severity levels may face higher medical expenditure risks, therefore granting them higher credit limits is reasonable.

[0127] To ensure that patients are assigned credit limits commensurate with their treatment costs based on the type of illness they suffer from, this embodiment utilizes the aforementioned evaluation system to more accurately match patients' medical needs, avoiding a "one-size-fits-all" approach to credit limit settings. For patients with high rehabilitation costs, the credit-based medical credit limit evaluation system can provide some financial support, alleviating their financial burden.

[0128] In order to ensure that the use of the credit limit is coordinated with medical insurance, the settlement credit limit in S5 of this embodiment... According to the formula:

[0129] ;

[0130] in For settlement amount, This refers to the reimbursement limit under medical insurance; when a patient does not use medical insurance, =0.

[0131] S6 patient's credit history for subsequent repayments Including the number of overdue payments Next time, overdue period Days, and on-time repayment rate Each evaluation factor is assigned a score, and the total score is calculated, based on:

[0132] ;

[0133] Credit limit Adjust to credit limit ,in For each overdue point, For points awarded if the deadline is one day overdue, The initial score is 10. For every 10% increase in on-time repayment rate, point, Where j is the weight, and the weight when the patient repays the loan during hospitalization is... The weighting for repayments made by patients outside of hospitalization is... On-time repayment rate According to the formula:

[0134] ;

[0135] in For loans that have matured and been repaid, These are loans that have already matured.

[0136] As a preferred technical solution in this embodiment, in the actual evaluation process, it can be... Set to 100 points. The assignment is divided into 5 parts, which can be... The assignment can be divided into 1, which can be... The assignment is divided into 2. This is achieved by setting the weight of repayments made during the patient's hospitalization period. The weighting for loan repayments made by patients outside of hospitalization is set as follows: Then, by calculating the patient's total credit history score and comparing it with the initial score, a credit limit adjustment ratio is determined, and the credit limit is adjusted accordingly. Adjust to credit limit This method provides a more direct view of a patient's credit history, minimizing the impact of late payments or other issues incurred during hospitalization on their credit limit. In the actual evaluation process... This evaluation method ensures that if a patient has fewer demerit points for overdue payments during hospitalization, the reduction in the credit limit will be smaller.

[0137] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A credit medical credit limit evaluation method based on decision boosting tree, comprising the following steps, characterized by, S1, a database DB for storing patient information is established, patients make outpatient registration and submit personal information to the database DB, the existing patient data in the database DB and the submitted patient personal information are matched, it is judged whether the patient has credit limit, if yes, S3 is entered, if not, S2 is entered; S2, evaluating a set of factors from the patient's social history to construct an initial creditworthiness evaluation system to grant the patient an initial credit line and proceed to S3; S3, the patient visits a doctor, the doctor diagnoses the patient's condition, and divides the patient into a short-term rehabilitation case and a long-term rehabilitation case according to the rehabilitation time of the condition S4, the patient is a long-term rehabilitation case S5, the patient is a short-term rehabilitation case S4, the patient is a long-term rehabilitation case S4, the patient is a long-term rehabilitation case S4, according to the long-term rehabilitation cases of patients to build a one-time quota adjustment evaluation system, through the one-time quota adjustment evaluation system to adjust the credit quota of patients to credit quota ; The long-term rehabilitation cases in S4 Different long-term rehabilitation disease types m are included, each long-term rehabilitation disease type m is divided into different severity grades y, different severity grades y correspond to different rehabilitation times k, and each different disease type m corresponds to different initial amounts Each different disease type m corresponds to different initial weights , define a mapping function : ； In order to avoid If the value is too large or too small, it affects the mapping function. The output value is truncated to make it fall within the range Between them, through the formula: ； wherein is a minimum output value of the mapping function, is a maximum output value of the mapping function, the mapping function is adjusted by a weight to obtain an adjusted weight : ； According to the adjusted weight and the initial credit limit for each m to derive the adjusted limit : ； to arrive at an adjusted amount and adjusting the initial credit amount to arrive at a secondary amount : ； wherein , , is a coefficient, is the shortest recovery time for a patient suffering from the mth disease category, is the longest recovery time for a patient suffering from the mth disease category; S5, the patient is treated according to the doctor's diagnosis, the treatment cost generated by the credit limit The credit limit is used to pay the treatment cost And the patient's treatment record is obtained; S6, after the treatment is completed, settlement is carried out, and during the settlement, the medical insurance reimbursement amount of the patient is used to realize the quota settlement, and the settlement amount is obtained ;​ S7, updating the one-time quota adjustment system according to the patient's treatment record and the actual settlement quota to establish a machine learning model, and update the one-time quota adjustment system. S8, the patient settles the amount making a repayment, according to the credit record of the patient's subsequent repayment establishing a secondary amount adjustment evaluation system to obtain the credit amount storing the amount information of the patient into a database DB, and updating the original patient information. 2.The credit medical limit evaluation method based on decision boosting tree according to claim 1, characterized in that, The set of social history evaluation factors in S2 , including x evaluation factors, each evaluation factor x having h rating levels, each evaluation factor x being assigned a weight w, and each rating level h being assigned a score j; through the formula: ； To obtain the patient's social history score F, the range of values ​​for the social history score F is determined. Divide into four sub-intervals from smallest to largest: ； wherein corresponding to different values of wherein corresponding to different values of , according to the formula: ； According to , a credit limit is derived. 3.The credit medical limit evaluation method based on decision boosting tree according to claim 1, characterized in that, The settlement amount in the S6 , according to the formula: ; wherein is the settlement amount, is the medical insurance reimbursement amount, when the patient does not use medical insurance, = 0.

4. The credit medical limit evaluation method based on the decision boosting tree according to claim 3, characterized in that, For patients with disease m, the actual settlement amount required for use is obtained by establishing a gradient boosting tree model to estimate the treatment cost of the patient to obtain an estimated cost , and by comparing the difference value between the estimated cost and the credit limit , the value range of is adjusted to build an initialization model: ； wherein is the number of samples, is the actual treatment cost of the i-th sample, is the initial estimate; Calculate the residual error: for each sample i, calculate the residual error between the current model prediction and the actual value: ； wherein, is the actual treatment cost for sample i, is the estimated cost for sample i, wherein is the feature vector for the ith sample; Fitting the residuals using a decision tree to obtain the prediction of the t-th tree ; ； where L is a loss function, is a tree learning rate, is the number of decision trees; Add a new decision tree to the existing prediction model to update the overall prediction to obtain the final prediction model: ； for predicting the treatment costs of new samples, according to: , a predicted treatment cost is derived from the deviation value of the credit limit , the value range of the mapping function is updated according to the size of > 0, is decreased, and when < 0, is increased, according to: ； to be adjusted, wherein is an adjustment step size determined according to the distribution and size of the errors, is an adjustment function.

5. The credit medical limit evaluation method based on the decision boosting tree according to claim 4, characterized in that, The tree is adjusted according to the number of diseases m suffered by the patient The learning rate of the tree is adjusted, specifically as follows: ； where k is the total number of different diseases the patient has, i is the ith disease the patient has, i = (1, 2,..., k), is the corresponding grade of the ith disease of the patient, and is a hyperparameter to control the degree of influence of the disease grade and the number of diseases on the learning rate, is an offset to adjust the base value of the learning rate.

6. The credit medical limit evaluation method based on the decision boosting tree according to claim 1, characterized in that, The S8 patient's subsequent repayment of the credit including the number of overdue times the number of times, the overdue time days, and the on-time repayment rate Each evaluation factor is scored and the total score is calculated, and according to: ； Adjusting the credit line to the credit line wherein is the score for one delinquency, is the score for one day delinquency, is the initial score; 10 For each 10% of the on-time payment rate , where j is the weight, the weight for the patient's inpatient period , is the weight for the patient's non-inpatient period.

7. The credit medical limit evaluation method based on the decision boosting tree according to claim 6, characterized in that, The repayment on time rate According to the formula: ； wherein is a paid off matured loan, is a matured loan.

Images