Personalized food-grade stomach-suppressing capsule usage recommendation system and method

Abstract

The present invention discloses a personalized food-grade stomach-sucking capsule use recommendation system and method, and relates to the technical field of stomach-sucking capsule recommendation system. The target value of each recommendation scheme in a similar user scenario is calculated through an objective function, and the target value is combined with a selection algorithm to delete some recommendation schemes. A mutation operation is performed for each recommended scheme that is retained to generate a corresponding variation scheme, and based on the comparison result between the target value of the recommended scheme and the target value of the variation scheme, it is determined whether to accept the variation scheme. According to the comparison result, a scheme set is constructed for all the retained recommended schemes. When the convergence condition is met, all scheme sets are output, all scheme sets are integrated into a total set, and the recommended scheme with the largest target value in the total set is selected and sent to the user. The recommendation system introduces a dynamic iteration mechanism, and through multiple rounds of calculation, screening and optimization, the recommended scheme is adjusted in real time to adapt to the changing needs of the user during use, and the adaptability of the recommendation system is enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of Zhanwei capsule recommendation systems, and in particular to a personalized food-grade Zhanwei capsule use recommendation system and method. Background Art

[0002] With the improvement of living standards and changes in dietary structure, the number of obese people in the world continues to increase. Obesity has become a global public health issue. Obesity not only affects appearance, but also leads to a significant increase in the risk of chronic diseases such as cardiovascular disease, diabetes, and hypertension. People are in urgent need of safe and effective weight management methods. Food-grade stomach-sucking capsules are a new type of health management product that aims to help consumers reduce food intake, control appetite, and achieve the purpose of weight management and health improvement by occupying a certain space in the stomach.

[0003] The prior art has the following deficiencies:

[0004] Most recommendation systems adopt static recommendation strategies, ignore dynamic feedback from users during use, lack a dynamic adjustment mechanism, are unable to balance the trade-offs between multiple goals (such as weight loss effect, safety, and user experience), and are difficult to adjust recommendation plans in real time. When user needs change or environmental conditions change, the recommendation results may no longer be applicable, thereby reducing dependence on the recommendation system and frequency of use.

[0005] Based on this, the present invention proposes a personalized food-grade stomach-sucking capsule usage recommendation system and method, introduces a dynamic iteration mechanism, and adjusts the recommendation scheme in real time through multiple rounds of calculation, screening and optimization to adapt to the changing needs of users during use and enhance the adaptability of the recommendation system. Summary of the invention

[0006] The purpose of the present invention is to provide a personalized food-grade stomach-replenishing capsule use recommendation system and method to address the deficiencies in the background technology.

[0007] In order to achieve the above object, the present invention provides the following technical solution: a personalized food-grade stomach-suppressing capsule use recommendation method, the recommendation method comprising the following steps:

[0008] The recommendation system randomly generates several recommendation solutions through a random generation tool, matching similar user scenarios based on the database;

[0009] The target value of each recommended solution in similar user scenarios is calculated through the objective function. The target value is combined with the selection algorithm to delete some recommended solutions. A mutation operation is performed for each recommended solution to generate a corresponding variant solution. Based on the comparison result between the target value of the recommended solution and the target value of the variant solution, it is determined whether to accept the variant solution. Based on the comparison result, all the recommended solutions retained are combined to form a solution set.

[0010] Repeat the target value calculation, deletion, and mutation steps for the recommended solutions in the solution set. When the convergence condition is met, output all solution sets, integrate all solution sets into a total set, select the recommended solution with the largest target value in the total set and send it to the user.

[0011] In a preferred embodiment, the recommended regimen includes the number of capsules to be used per day, the time of use, and the frequency of use.

[0012] In a preferred embodiment, obtaining current user data and matching similar user scenarios based on a database includes the following steps:

[0013] Extract the user's health characteristics and usage needs, and provide input data for similar user matching. The input data includes the feature vector of the current user.

[0014] Standardize historical user data and define the historical user set as: , where Represents the number of historical users, and the feature normalization formula is: , where For the User in The standard value of the feature, For the User in The original value of the feature, For the The mean of the features, For the The standard deviation of the feature;

[0015] Calculate the distance between the current user and the historical user, and The distance between each user and the current user is converted into a similarity score, expressed as: , where For the The similarity score between the user and the current user, For the The distance between the user and the current user is calculated, and the users in the database whose similarity scores are greater than or equal to the score threshold are selected as the similar user scenarios of the current user. If the number of users whose similarity scores are greater than or equal to the score threshold is 0, the user with the largest similarity score in the database is selected as the similar user scenario of the current user.

[0016] In a preferred embodiment, the distance between the current user and the historical user is calculated using the expression:

[0017] , where For the The distance between each user and the current user, is the number of eigenvectors, For the User in The standard value of the feature, For the current user in The standard value of a feature.

[0018] In a preferred embodiment, the target value of each recommendation scheme in a similar user scenario is calculated by an objective function, including the following steps:

[0019] The target value of each recommended solution in similar user scenarios is calculated through the objective function, and the expression is: , where is the target value, is a positive return factor, is a negative impact factor. , are the adjustment coefficients of the positive benefit factor and the negative impact factor, respectively, and , Both are greater than 0.

[0020] In a preferred embodiment, combining the target value with the selection algorithm to delete some recommended solutions includes the following steps:

[0021] After obtaining the target value of each recommended solution, sum up the target values ​​of all recommended solutions to obtain the total target value, divide the target value by the total target value to obtain the selection probability of each recommended solution, map the selection probability to the sector area of ​​the virtual roulette, start the virtual roulette to rotate, and when the virtual roulette stops, point the pointer to the recommended solution corresponding to the sector area and select it, and repeatedly start the virtual roulette to perform the selection operation. When the number of selected recommended solutions is equal to the quantity threshold, the selected recommended solutions are retained and the unselected recommended solutions are deleted.

[0022] In a preferred embodiment, judging whether to accept the variant solution based on the comparison result between the target value of the recommended solution and the target value of the variant solution includes the following steps:

[0023] After obtaining the target value of the recommended solution and the target value of the variant solution, if , judge that the variation scheme is not accepted, if , determine whether to accept the variant solution, replace the recommended solution with the variant solution, and after completing the comparison between the target values ​​of all recommended solutions and the target values ​​of the variant solutions, construct a solution set for all retained recommended solutions.

[0024] A personalized food-grade stomach-sucking capsule usage recommendation system, including a solution generation module, a collection construction module, and a recommendation module;

[0025] Solution generation module: randomly generates several recommended solutions through random generation tools, obtains current user data, and matches similar user scenarios based on the database;

[0026] Set construction module: Calculate the target value of each recommended solution in similar user scenarios through the objective function, combine the target value with the selection algorithm to delete some recommended solutions, perform mutation operations on each recommended solution to generate corresponding variant solutions; determine whether to accept the variant solution based on the comparison results between the target value of the recommended solution and the target value of the variant solution, and construct a solution set for all the retained recommended solutions according to the comparison results;

[0027] Recommendation module: Repeat the target value calculation, deletion, and mutation steps for the recommended solutions in the solution set. When the convergence conditions are met, output all solution sets, integrate all solution sets into a total set, select the recommended solution with the largest target value in the total set and send it to the user.

[0028] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0029] The present invention calculates the target value of each recommended solution in a similar user scenario through an objective function, combines the target value with the selection algorithm to delete some recommended solutions, performs mutation operations for each recommended solution to generate corresponding variant solutions, and determines whether to accept the variant solution based on the comparison result between the target value of the recommended solution and the target value of the variant solution. According to the comparison result, all the recommended solutions retained are constructed into a solution set. When the convergence condition is met, all solution sets are output, all solution sets are integrated into a total set, and the recommended solution with the largest target value in the total set is selected and sent to the user. The recommendation system introduces a dynamic iteration mechanism, and through multiple rounds of calculation, screening and optimization, the recommended solution is adjusted in real time to adapt to the changing needs of users during use, thereby enhancing the adaptability of the recommendation system. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0031] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0032] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] Example 1: Please refer to Figure 1 As shown, the recommended method for using the personalized food-grade stomach-suppressing capsule in this embodiment includes the following steps:

[0034] The recommendation system randomly generates several recommendation plans through a random generation tool. The content of the recommendation plan includes the daily number of capsules used, usage time and usage frequency, obtains current user data, matches similar user scenarios based on the database, calculates the target value of each recommendation plan in similar user scenarios through the objective function, combines the target value with the selection algorithm to delete some recommendation plans, performs mutation operations for each retained recommendation plan to generate corresponding mutation plans, and determines whether to accept the mutation plan based on the comparison result between the target value of the recommended plan and the target value of the mutation plan. According to the comparison result, a solution set is constructed for all retained recommendation plans, and the target value calculation, deletion and mutation steps of the recommended plans in the solution set are repeated. When the convergence condition is met, all solution sets are output, all solution sets are integrated into a total set, and the recommendation plan with the largest target value in the total set is selected and sent to the user.

[0035] This application calculates the target value of each recommended solution in similar user scenarios through an objective function, combines the target value with the selection algorithm to delete some recommended solutions, performs mutation operations for each recommended solution to generate corresponding variant solutions, and determines whether to accept the variant solution based on the comparison results between the target value of the recommended solution and the target value of the variant solution. According to the comparison results, all the retained recommended solutions are constructed into a solution set. When the convergence conditions are met, all solution sets are output, all solution sets are integrated into a total set, and the recommended solution with the largest target value in the total set is selected and sent to the user. The recommendation system introduces a dynamic iteration mechanism, which adjusts the recommended solution in real time through multiple rounds of calculation, screening and optimization to adapt to the changing needs of users during use and enhance the adaptability of the recommendation system.

[0036] In this application, the application scenarios of the recommended method can be widely applied to the following situations, especially in the fields of health management, weight loss and personalized nutritional supplementation. The following are several specific application scenario examples:

[0037] 1) Personalized weight loss plan recommendation:

[0038] Users hope to use food-grade Zhanwei capsules to help control appetite and achieve weight loss goals. Each user's physique, eating habits, exercise volume, and weight loss goals are different, so personalized capsule usage recommendations are needed.

[0039] The recommendation system will randomly generate multiple capsule usage plans, each of which includes the number of capsules to be used per day, the time of use, and the frequency of use. Based on the user's health data, weight, eating habits and other information, the system uses an objective function to calculate the effect of each recommended plan in similar user scenarios, such as weight loss effect, appetite control, etc. Through the calculation and selection algorithm of the target value, the plans with poor effects are deleted, the optimized plans are retained and mutated to generate new mutated plans. After multiple rounds of iteration and optimization, the final output is the weight loss plan that best meets the user's needs, and the most effective capsule usage frequency and time are recommended to achieve the maximum weight loss effect.

[0040] Example: A user is overweight and has a high appetite. Through the personalized recommendation system, the final recommended capsule usage plan is to use 2 capsules per day, 30 minutes before meals and before bedtime, once a day.

[0041] 2) Optimizing the use of medicines for special health needs

[0042] Some patients may need to use Zanwei Capsules to assist with treatment, such as by regulating the stomach to control appetite, in conjunction with the treatment of certain digestive disorders, or as a way to control diet. Personalized recommendations can ensure the optimal use of medication to support the effectiveness of the treatment plan.

[0043] The system generates different recommended plans for capsule use based on the patient's physical condition, medication history, and treatment plan. The system calculates the target value of each plan in the patient's current treatment plan and analyzes the impact of each recommended plan on disease management. Through multiple rounds of variation and target value comparison, the system finds the optimal plan that suits the patient's health status and adjusts the time and frequency of capsule use to support the patient's recovery or treatment effect.

[0044] Example: A diabetic patient wants to control his appetite to avoid overeating. Through the recommendation system, the final recommended solution may be to use 1 capsule in the morning and 1 capsule at noon every day to help reduce excessive appetite before meals and lose weight with diet control.

[0045] 3) Personalized nutrition plan for health management and weight control

[0046] Scenario description: Some people may want to control their weight through a personalized nutrition plan, especially when they need to use food-grade stomach-suppressing capsules for a long time. Personalized recommendations can help users control their food intake and achieve weight management while maintaining good health.

[0047] The system collects the user's physical data, including weight, height, age, gender, amount of exercise, etc., and generates different recommended plans for capsule use. The system calculates the target value for each plan based on the user's health goals (for example, weight loss or weight maintenance) and optimizes it to ensure that the recommended plan best meets the user's health management goals. Through multiple rounds of iterative optimization, the system provides the most suitable number, time, and frequency of capsule use for the user to ensure the best weight management effect.

[0048] Example: A user who is losing weight wants to avoid overeating by regulating appetite. The recommendation system suggests that the user take 1 capsule each morning and evening and follow a low-calorie diet, with the goal of controlling his daily calorie intake.

[0049] 4) Health monitoring and behavior change incentive system

[0050] This recommendation method can also be used in health monitoring and behavior change systems, for example, by combining it with smart hardware (such as smart watches, health trackers, etc.) to adjust the capsule's recommendation plan based on real-time data (such as weight, exercise volume, appetite control, etc.).

[0051] The system obtains real-time user data from smart hardware and health platforms, such as weight changes, diet, and exercise. Based on the user's health goals (such as losing weight, staying healthy, etc.), the system calculates the effect of each recommended solution through an objective function. As the user's behavior and health data change, the system will dynamically adjust the use time, frequency, and quantity of the recommended solution to maintain a timely response to user needs.

[0052] Example: A user tracks his daily activities through a smart bracelet. The system detects that the user has reduced his exercise and gained weight over a period of time. Therefore, the recommendation system dynamically adjusts the use of capsules and increases the frequency of use of capsules to help curb excessive eating and promote weight management.

[0053] 5) Appetite control management for high-risk groups

[0054] Scenario description:

[0055] For some high-risk groups suffering from obesity, metabolic disorders and other problems, the personalized recommendation system can provide a scientific and reasonable appetite control plan to help them take food-grade Zhanwei capsules at the appropriate time and frequency to support overall health management.

[0056] The system first obtains the user's health history data, such as obesity level, blood sugar level, eating habits, etc. Based on this data, it generates different recommended plans and calculates and compares their effects based on the target value. During the iteration process, the plan will be continuously adjusted based on the user's feedback and health data, and finally the most suitable personalized appetite control plan will be given.

[0057] Example: A user with obesity wants to control appetite and reduce overeating habits by taking a stomach-warming capsule. The recommendation system may recommend taking 1 capsule twice a day, 30 minutes before the main meal, and combining it with appropriate exercise to improve the effect.

[0058] Personalized food-grade Zhanwei capsule usage recommendation method can play a role in multiple scenarios, especially in the fields of weight loss, health management, disease treatment and appetite control. Through intelligent data analysis and personalized recommendation algorithms, it can help users customize the most suitable capsule usage plan according to their personal physical condition, health goals and lifestyle habits, so as to achieve the best health management effect.

[0059] Example 2: The recommendation system randomly generates a number of recommendation plans through a random generation tool, and the recommendation plans include the number of capsules to be used per day, the time of use, and the frequency of use, including the following steps:

[0060] Objective: Define the range and randomization rules for generating recommendations.

[0061] Define variable ranges: Determine the core variables and their value ranges in the recommended solution, for example:

[0062] Daily capsule intake: minimum and maximum values ​​(e.g. 1-6 capsules).

[0063] When to use: Time of day (e.g. before breakfast, after lunch, after dinner).

[0064] Frequency of use: The number of times taken in a day (e.g. 1 time, 2 times, 3 times).

[0065] Set the random generation rules:

[0066] Use random generation tools to randomize variable combinations to ensure the diversity of recommended solutions.

[0067] Constraints can be introduced, for example: the total amount of capsules per day does not exceed a maximum value, the frequency of use is limited within a specific time period, etc.

[0068] Objective: Randomly generate multiple initial recommendation solutions.

[0069] Number of generated solutions: Determine the number of recommended solutions (e.g. 100 solutions) based on system performance and user scenario requirements.

[0070] Random combination generation: Randomly select the number of capsules to be used per day. Randomly assign specific time periods for use (e.g., breakfast, lunch, dinner). Randomly determine the daily frequency of use and assign specific times for use.

[0071] Generate a proposal format: Each recommended proposal includes the following:

[0072] -Daily use quantity: 4 pieces;

[0073] -Use time: 2 capsules before breakfast, 1 capsule after lunch, and 1 capsule after dinner;

[0074] - Frequency of use: 3 times / day.

[0075] Obtain the current user data and match similar user scenarios based on the database, including the following steps:

[0076] Extract the user's health characteristics and usage needs, and provide input data for similar user matching. The input data includes the feature vector of the current user:

[0077] Physiological characteristics: weight, height, BMI, basal metabolic rate, etc.

[0078] Health status: blood sugar level, allergy information, underlying disease conditions, etc.

[0079] Behavioral characteristics: eating habits, amount of exercise, work and rest patterns, etc.

[0080] Usage goals: weight loss goals, daily calorie intake goals, etc.

[0081] Standardize historical user data to prepare for similarity calculation and define the historical user set as: , where Indicates the number of historical users, each With current user characteristics The feature dimension is consistent, and the feature standardization formula is: , where For the User in The standard value of the feature, For the User in The original value of the feature, For the The mean of the features, For the The standard deviation of the feature;

[0082] Calculate the distance between the current user and the historical users, select the most similar users from the database, and calculate the distance using the Euclidean distance. The expression is: , where For the The distance between each user and the current user, is the number of eigenvectors, For the User in The standard value of the feature, For the current user in Standard value on a feature;

[0083] The first The distance between each user and the current user is converted into a similarity score, expressed as:

[0084] , where For the The similarity score between the user and the current user, For the The distance between the user and the current user is calculated, and the users in the database whose similarity scores are greater than or equal to the score threshold are selected as the similar user scenarios of the current user. If the number of users whose similarity scores are greater than or equal to the score threshold is 0, the user with the largest similarity score in the database is selected as the similar user scenario of the current user.

[0085] The target value of each recommended solution in a similar user scenario is calculated through the objective function, including the following steps:

[0086] The target value of each recommended solution in similar user scenarios is calculated through the objective function, and the expression is: , where is the target value, is a positive return factor, is a negative impact factor. , are the adjustment coefficients of the positive benefit factor and the negative impact factor, respectively, and , Both are greater than 0.

[0087] The calculation logic of the positive benefit factor is: since users with similarity scores greater than or equal to the score threshold in the database are selected as similar user scenarios for the current user, if the number of users with similarity scores greater than or equal to the score threshold is 0, the user with the largest similarity score in the database is selected as the similar user scenario for the current user. Therefore, there is at least one user in the similar user scenario. When there is one user in the similar user scenario, the positive benefit factor is the shock absorption effect of the user. When there are two or more users in the similar user scenario, the mean shock absorption effect and the standard deviation of the shock absorption effect of the similar user scenario are calculated first, and the positive benefit factor is calculated. The expression is: , where is a positive return factor, is the average value of shock absorption effect, is the standard deviation of the shock-absorbing effect. The larger the positive benefit factor is, the better the shock-absorbing effect of the recommended plan on the user. The weight difference is obtained by subtracting the final weight from the user's initial weight, and the weight difference is divided by the duration of taking the medicine to obtain the weight loss effect.

[0088] The calculation logic of the negative impact factor is as follows: obtain the side effect incidence and the abandonment rate of the recommended regimen, normalize the side effect incidence and the abandonment rate of the regimen, map the value range of the side effect incidence and the abandonment rate to [0,1], obtain the normalized value of the side effect incidence and the normalized value of the abandonment rate of the regimen, sum the normalized value of the side effect incidence and the normalized value of the abandonment rate of the regimen to obtain the negative impact factor. The larger the negative impact factor, the more obvious the side effects of the recommended regimen or the higher the abandonment rate. The side effect incidence is obtained by dividing the number of side effect occurrences by the duration of medication, and the abandonment rate is obtained by dividing the number of regimen abandonments by the duration of medication.

[0089] The target value is combined with the selection algorithm to delete some recommended solutions, and a mutation operation is performed for each recommended solution to generate a corresponding mutation solution, including the following steps:

[0090] After obtaining the target value of each recommended solution, sum up the target values ​​of all recommended solutions to obtain the total target value, divide the target value by the total target value to obtain the selection probability of each recommended solution, and map the selection probability to the sector area of ​​the virtual roulette. The larger the selection probability, the larger the corresponding sector area. Start the virtual roulette to rotate. When the virtual roulette stops, point the pointer to the recommended solution corresponding to the sector area and select it. Repeat the virtual roulette to perform the selection operation. When the number of selected recommended solutions is equal to the quantity threshold, the selected recommended solutions are retained and the unselected recommended solutions are deleted. Through this selection algorithm, the selection probability of recommended solutions with larger target values ​​can be improved, and after increasing the randomness, it is beneficial to improve the robustness and adaptability of the recommendation system.

[0091] Perform mutation operations on all retained recommendation solutions. The mutation operation involves modifying the following parameters of the recommendation solution:

[0092] 1. Daily capsule usage quantity:

[0093] How it works: Vary by increasing or decreasing the number of capsules taken daily.

[0094] Range: A reasonable range can be set based on the individual needs of the user. For example, if the recommended regimen was originally 2 capsules per day, it can be varied to any value between 1 and 3 capsules per day.

[0095] 2. Usage time:

[0096] How to use: Change the time of use of the capsules, or schedule them for different times of the day.

[0097] Scope: For example, changing the original recommendation of "take in the morning" to "take before lunch" or "take at night before bed" can be adjusted to different time periods of the day.

[0098] 3. Frequency of use:

[0099] How to operate: Adjust the number of times you take the medicine per day, or change the frequency of taking it.

[0100] Range: For example, change "once a day" to "twice a day" or "take every other day."

[0101] According to the mutation method, perform the corresponding mutation operation on each selected recommendation solution. The specific operation methods include:

[0102] 1. Variation in the number of capsules used per day:

[0103] By increasing or decreasing a certain number (for example, originally 2 pills, changed to 2±1 pills), we can ensure that the changed value still falls within the health recommendation range.

[0104] 2. Use time variation: Randomly select different time periods of the day for adjustment, for example: adjust from 7 am to 12 noon, or from 9 am to 8 pm.

[0105] 3. Frequency variation: Randomly adjust the frequency of use to ensure that the number of daily doses is within a reasonable range. For example, the original plan is once a day, and the modified plan may be twice a day or once every two days.

[0106] The mutation operation should ensure that the change of each parameter is reasonable and does not negatively affect the user's health.

[0107] Based on the comparison result between the target value of the recommended solution and the target value of the variant solution, it is determined whether to accept the variant solution, and a solution set is constructed by combining all the retained recommended solutions according to the comparison result, including the following steps:

[0108] After obtaining the target value of the recommended solution and the target value of the variant solution, if , judge that the variation scheme is not accepted, if , determine whether to accept the variant solution, replace the recommended solution with the variant solution, and after completing the comparison between the target values ​​of all recommended solutions and the target values ​​of the variant solutions, construct a solution set for all retained recommended solutions.

[0109] Repeat the target value calculation, deletion, and mutation steps for the recommended solutions in the solution set. When the convergence condition is met, output all solution sets, integrate all solution sets into a total set, select the recommended solution with the largest target value in the total set and send it to the user, including the following steps:

[0110] During the iteration process, the system needs to determine whether the convergence conditions are met. Common convergence conditions include:

[0111] Maximum number of iterations: When the number of iterations equals the preset number threshold, that is, reaches the predetermined upper limit, the iteration stops.

[0112] The target value change is less than the change threshold: When the improvement of the target value is less than the preset threshold in several consecutive iterations, the system is considered to have converged.

[0113] When the convergence condition is met, all the retained recommended solutions are output. These solutions are the optimal solution set after multiple rounds of iterative optimization. Select the recommended solution with the highest target value from the final solution set. This solution will be the recommended solution that best meets user needs and has the best effect. This can be achieved by following the following steps:

[0114] Sort all the plans according to their target values ​​from large to small, and select the plan with the highest target value.

[0115] , where Represents the number of recommended solutions in the total set, and selects the recommended solution with the highest target value from the sorted solutions, that is, , and take it as the final recommended solution.

[0116] Finally, the recommended solution with the highest target value is sent to the user to complete the recommendation process. The recommended solution will be optimized based on the user's needs and health data to maximize the weight loss effect and user satisfaction. By repeatedly calculating the target value, deleting solutions with poor target values, performing mutation operations, and judging whether to accept the mutated solution, a set of solutions that have undergone multiple rounds of iterative optimization is finally formed. When the convergence conditions are met, the system outputs the recommended solution with the highest target value and sends it to the user. This process ensures the personalization, accuracy, and efficiency of the recommended solution, and maximizes user needs.

[0117] Embodiment 3: The personalized food-grade stomach-suppressing capsule usage recommendation system described in this embodiment includes a solution generation module, a set construction module, and a recommendation module;

[0118] Solution generation module: randomly generates several recommended solutions through random generation tools. The recommended solutions include the number of capsules used per day, the time of use, and the frequency of use. The current user data is obtained, and similar user scenarios are matched based on the database. The recommended solutions and similar user scenarios are sent to the collection construction module.

[0119] Set construction module: Calculate the target value of each recommended solution in similar user scenarios through the objective function, combine the target value with the selection algorithm to delete some recommended solutions, perform mutation operations on each recommended solution to generate corresponding variant solutions; determine whether to accept the variant solution based on the comparison result between the target value of the recommended solution and the target value of the variant solution, and construct a solution set for all the retained recommended solutions according to the comparison result, and send the solution set to the recommendation module;

[0120] Recommendation module: Repeat the target value calculation, deletion, and mutation steps for the recommended solutions in the solution set. When the convergence conditions are met, output all solution sets, integrate all solution sets into a total set, select the recommended solution with the largest target value in the total set and send it to the user.

[0121] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0122] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0123] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0124] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process of the above-described system, device and unit can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0125] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. Recommended method for using personalized food-grade stomach-suppressing capsules, characterized in that: The recommended method comprises the following steps: The recommendation system randomly generates several recommendation solutions through a random generation tool, matching similar user scenarios based on the database; The target value of each recommended solution in similar user scenarios is calculated through the objective function, and the expression is: , where is the target value, is a positive return factor, is a negative impact factor. , are the adjustment coefficients of the positive benefit factor and the negative impact factor, respectively, and , All are greater than 0, and the calculation expression of the positive return factor is: , where is a positive return factor, is the mean value of weight loss effect, The standard deviation of the weight loss effect was calculated by summing the normalized values ​​of the side effect incidence rate and the normalized values ​​of the regimen abandonment rate to obtain the negative impact factor; Combine the target value with the selection algorithm to delete some recommended solutions, including: After obtaining the target value of each recommended solution, sum up the target values ​​of all recommended solutions to obtain the total target value, divide the target value by the total target value to obtain the selection probability of each recommended solution, map the selection probability to the sector area of ​​the virtual roulette, start the rotation of the virtual roulette, and when the virtual roulette stops, point the pointer to the recommended solution corresponding to the sector area and select it, and repeatedly start the virtual roulette to perform the selection operation. When the number of selected recommended solutions equals the quantity threshold, the selected recommended solutions are retained, and the unselected recommended solutions are deleted; Perform mutation operations on each recommended solution to generate corresponding variant solutions, and determine whether to accept the variant solution based on the comparison between the target value of the recommended solution and the target value of the variant solution. Construct a solution set for all the recommended solutions retained based on the comparison results. Repeat the target value calculation, deletion, and mutation steps for the recommended solutions in the solution set. When the convergence condition is met, output all solution sets, integrate all solution sets into a total set, select the recommended solution with the largest target value in the total set and send it to the user.

2. The personalized food-grade stomach-replenishing capsule use recommendation method according to claim 1, characterized in that: The recommended regimen includes the number of capsules to be used per day, the time of use, and the frequency of use.

3. The personalized food-grade stomach-replenishing capsule use recommendation method according to claim 2, characterized in that: Obtain the current user data and match similar user scenarios based on the database, including the following steps: Extract the user's health characteristics and usage needs, and provide input data for similar user matching. The input data includes the feature vector of the current user. Standardize historical user data and define the historical user set as: , where Represents the number of historical users, and the feature normalization formula is: , where For the User in The standard value of the feature, For the User in The original value of the feature, For the The mean of the features, For the The standard deviation of the feature; Calculate the distance between the current user and the historical user, and The distance between each user and the current user is converted into a similarity score, expressed as: , where For the The similarity score between the user and the current user, For the The distance between the user and the current user is calculated, and the users in the database whose similarity scores are greater than or equal to the score threshold are selected as the similar user scenarios of the current user. If the number of users whose similarity scores are greater than or equal to the score threshold is 0, the user with the largest similarity score in the database is selected as the similar user scenario of the current user.

4. The personalized food-grade stomach-replenishing capsule use recommendation method according to claim 3, characterized in that: Calculate the distance between the current user and the historical user. The expression is: , where For the The distance between each user and the current user, is the number of eigenvectors, For the User in The standard value of the feature, For the current user in The standard value of a feature.

5. The personalized food-grade stomach-replenishing capsule use recommendation method according to claim 4, characterized in that: Based on the comparison result between the target value of the recommended solution and the target value of the variant solution, whether to accept the variant solution is determined, including the following steps: After obtaining the target value of the recommended solution and the target value of the variant solution, if , judge that the variation scheme is not accepted, if , determine whether to accept the variant solution, replace the recommended solution with the variant solution, and after completing the comparison between the target values ​​of all recommended solutions and the target values ​​of the variant solutions, construct a solution set for all retained recommended solutions.

6. A personalized food-grade stomach-suppressing capsule use recommendation system, used to implement the recommendation method according to any one of claims 1 to 5, characterized in that: It includes solution generation module, set building module and recommendation module; Solution generation module: randomly generates several recommended solutions through random generation tools, obtains current user data, and matches similar user scenarios based on the database; Set building module: Calculate the target value of each recommendation scheme in similar user scenarios through the objective function, combine the target value with the selection algorithm to delete some recommendation schemes, and perform mutation operations for each recommended scheme that is retained to generate corresponding mutation schemes; Based on the comparison results between the target values ​​of the recommended solution and the target values ​​of the variant solution, determine whether to accept the variant solution, and construct a solution set for all the retained recommended solutions according to the comparison results; Recommendation module: Repeat the target value calculation, deletion, and mutation steps for the recommended solutions in the solution set. When the convergence conditions are met, output all solution sets, integrate all solution sets into a total set, select the recommended solution with the largest target value in the total set and send it to the user.

Images