Individual interpersonal contact number prediction method, system and device and storage medium

Abstract

The invention relates to an individual interpersonal contact number prediction method, system and device and a storage medium. The method comprises the following steps: obtaining questionnaire survey data of individual resident travel activities and interpersonal contact; extracting individual attribute features and travel activity features of individual residents from the questionnaire survey data; constructing and training an individual interpersonal contact mode prediction model by applying a machine learning algorithm based on the demographic features and the travel activity features; and performing individual interpersonal contact number prediction by using the trained individual interpersonal contact mode prediction model, and outputting an individual interpersonal contact number prediction value. According to the method, the dynamic travel activity features and the static demographic attributes are fused, and machine learning modeling is utilized, so that refined individual behaviors can be captured, the number of individual contact persons can be more reliably predicted, and the assessment accuracy of the disease transmission risk is improved.

Description

Technical Field

[0001] This application belongs to the fields of public health and spatiotemporal data mining technology, and specifically relates to a method, system, device and storage medium for predicting the number of interpersonal contacts of an individual. Background Technology

[0002] The number of individual interpersonal contacts is a key parameter in many infectious disease transmission models. In the past few decades, the public health field has mainly obtained data on the number of individual interpersonal contacts through direct observation, questionnaires, and wearing sensors. These methods are costly and make it difficult to collect sufficiently representative samples covering different regions. Therefore, recent studies (after 2022) have begun to introduce travel activity characteristics to predict the number of individual interpersonal contacts. Early studies on interpersonal contact pattern prediction (before 2021) did not explore the relationship between mobility and contact patterns and often relied on demographic characteristics such as age, gender, and income. For example, Prem et al. [Prem K, Cook AR, Jit M. Projecting social contactmatrices in 152 countries using contact surveys and demographic data. PlosComput Biol 2017;13(9):e1005697.] constructed a Bayesian hierarchical model based on existing POLYMOD survey data and combined with multi-source information such as population census, labor participation rate, and school enrollment rate, and then predicted the social contact matrix of 152 countries. This method integrates contact survey data with demographic characteristics of different countries, enabling reasonable estimates of the number of interpersonal contacts among age groups in various settings such as families, workplaces, and schools in countries lacking field survey data. This prediction approach based on static demographic characteristics effectively compensates for the lack of contact data in many countries, while also highlighting the crucial role of individual social behavior and population structure in infectious disease transmission modeling.

[0003] However, research shows that the number of interpersonal contacts an individual has is not only influenced by static demographic characteristics such as age, gender, and occupation, but is also closely related to actual travel activity patterns. Therefore, scholars have begun to introduce human mobility data to more effectively and indirectly predict interpersonal contact patterns. Based on differences in data granularity and modeling methods, existing research can be broadly divided into two categories: I. Methods based on regionally aggregated travel activity characteristics; these methods are mostly driven by data availability, using regionally aggregated population flow data (e.g., population inflow / outflow) to predict the number of interpersonal contacts an individual will have. Their advantage lies in the ease of data acquisition and broad coverage, making them convenient to integrate into macro-epidemiological models. However, the difficulty in capturing refined individual behavior limits the model's effectiveness. II. Methods Based on Individual Travel Activity Characteristics; As a cutting-edge exploration, a few studies have attempted to introduce extremely high-precision individual mobility data (based on GPS positioning, with an accuracy within 5 meters) for direct contact prediction. This method, by incorporating individual travel activity characteristics, can capture behavioral patterns that dynamically change over time and space, thereby more accurately reconstructing contact characteristics in the real world and improving the predictive ability of individual interpersonal contact patterns. However, because this method relies entirely on extremely high-precision individual mobility data for direct contact prediction, it raises data privacy issues and lacks verification with real contact data, leading to doubts about the reliability of the results. Summary of the Invention

[0004] This application provides a method, system, device, and storage medium for predicting the number of interpersonal contacts of an individual, aiming to at least partially solve one of the aforementioned technical problems in the prior art.

[0005] To address the above problems, this application provides the following technical solution: A method for predicting the number of interpersonal contacts an individual may have includes: Obtain questionnaire survey data on individual residents' travel activities and interpersonal contacts; Individual resident attributes and travel activity characteristics were extracted from the questionnaire survey data. Based on the aforementioned demographic and travel activity characteristics, a machine learning algorithm is applied to construct and train an individual interpersonal contact pattern prediction model. The trained individual interpersonal contact pattern prediction model is used to predict the number of interpersonal contacts for an individual, and the predicted value of the number of interpersonal contacts for an individual is output.

[0006] The technical solution adopted in this application embodiment further includes: after obtaining the questionnaire survey data on individual residents' travel activities and interpersonal contacts, it also includes: The questionnaire survey data is preprocessed, including anomaly detection and invalid data removal.

[0007] The technical solution adopted in this application embodiment further includes: preprocessing the questionnaire survey data, including: The questionnaire data is reviewed, and invalid questionnaire data that contains logical inconsistencies, highly consistent options, or obvious patterns of responses is filtered out and removed. Outlier detection was performed on the questionnaire data based on the Raida criterion, and questionnaire data containing outliers were removed.

[0008] The technical solution adopted in this application embodiment further includes: the individual attribute characteristics include the individual resident's gender, age, and monthly income; the travel activity characteristics include turning radius, furthest distance from home, mobility entropy, frequency of daily activity types, student / employment status identifier, attendance status identifier, commuting distance, frequency of other daily activities, and status identifier of other daily activities; wherein, the turning radius is used to measure the individual's activity range, the furthest distance from home refers to the maximum geographical distance between all individual travel destinations or trajectory points and the home address within a day, the mobility entropy is an indicator that measures the complexity and diversity of individual travel behavior, and the daily activity characteristics include: turning radius, furthest distance from home, mobility entropy, and other daily activity characteristics. Activity frequency refers to the total number of all types of activities that an individual engages in on the statistical day. The "Study / Employment Status" identifier is used to identify whether the individual is currently a student or an employee. The "Daily Attendance Status" identifier is used to identify whether the individual attended school or work on the statistical day. The "Daily Commuting Distance" refers to the commuting distance an individual travels to school or work on the statistical day. The "Daily Other Activity Frequency" refers to the total number of other activities an individual engages in on the statistical day besides attending school or work. The "Daily Other Activity Status" identifier is used to identify whether the individual engages in other activities besides attending school or work on the statistical day.

[0009] The technical solution adopted in this application embodiment also includes: the construction and training of an individual interpersonal contact pattern prediction model based on the demographic characteristics and travel activity characteristics using machine learning algorithms, specifically: The questionnaire data is divided into training and testing sets, and the individual attribute features and travel activity features are used as input variables to train a machine learning model. The machine learning model includes Lasso regression model, ElasticNet regression model, CatBoost model, LightGBM model, XGBoost model, random forest, and multilayer perceptron.

[0010] The technical solution adopted in this application embodiment further includes: the step of constructing and training an individual interpersonal contact pattern prediction model based on the demographic characteristics and travel activity characteristics using machine learning algorithms, which further includes: During model training, the default parameters of each machine learning model are first used as a baseline, followed by hyperparameter optimization using a random cross-reference method. The optimization process of the random cross-reference method includes: Parameter space: Specifies a probability distribution for each hyperparameter; Random sampling: randomly selecting from the parameter space Group hyperparameter configuration, among which The number of iterations is preset. Evaluation and selection: Evaluate the model performance of each hyperparameter configuration on the validation set and select the hyperparameter configuration with the best performance.

[0011] The technical solution adopted in this application embodiment further includes: the step of constructing and training an individual interpersonal contact pattern prediction model based on the demographic characteristics and travel activity characteristics using machine learning algorithms, which further includes: The coefficient of determination, mean absolute percentage error, mean absolute error, and root mean square error are used as evaluation indicators for the predictive performance and error of the individual interpersonal contact pattern prediction model. The coefficient of determination represents the proportion of data variation explained by the model, the mean absolute percentage error is used to measure the error between the model's predicted value and the actual value, the mean absolute error reflects the average relative deviation between the predicted value and the actual value in the form of a percentage, and the root mean square error represents the average value of the absolute difference between the predicted value and the actual value.

[0012] Another technical solution adopted in this application embodiment is: an individual interpersonal contact number prediction system, including: Data acquisition module: used to acquire questionnaire survey data on individual residents' travel activities and interpersonal contacts; Feature extraction module: used to extract individual attribute features and travel activity features of individual residents from the questionnaire survey data; Model building module: used to build and train an individual interpersonal contact pattern prediction model based on the demographic characteristics and travel activity characteristics, using machine learning algorithms; Interpersonal contact prediction module: This module is used to predict the number of interpersonal contacts an individual will have using a pre-trained individual interpersonal contact pattern prediction model, and outputs the predicted number of interpersonal contacts.

[0013] Another technical solution adopted in this application embodiment is: a device, the device including a processor and a memory coupled to the processor, wherein, The memory stores program instructions for implementing the method for predicting the number of interpersonal contacts of an individual; The processor is used to execute the program instructions stored in the memory to control the method for predicting the number of interpersonal contacts of an individual.

[0014] Another technical solution adopted in this application embodiment is: a storage medium storing processor-executable program instructions, the program instructions being used to execute the individual interpersonal contact number prediction method.

[0015] Compared to existing technologies, the beneficial effects of the embodiments of this application are as follows: The method, system, device, and storage medium for predicting the number of individual interpersonal contacts in the embodiments of this application extract individual attribute features and travel activity features from questionnaire survey data, and then use machine learning algorithms to construct an individual interpersonal contact pattern prediction model, ultimately outputting the number of individual interpersonal contacts, thus achieving accurate prediction from individual characteristics to contact patterns. By integrating dynamic travel activity features with static demographic attributes and using machine learning modeling, this application can capture refined individual behavior and achieve more reliable prediction of the number of individual contacts. This not only provides high-precision data support for infectious disease transmission models, improves the accuracy of disease transmission risk assessment, and enhances the pertinence and effectiveness of epidemic prevention and control strategies, but also provides transferable practical examples for similar studies in other cities or regions. Attached Figure Description

[0016] Figure 1 This is a schematic flowchart of the method for predicting the number of interpersonal contacts of an individual according to an embodiment of this application; Figure 2 This is a schematic diagram of the structure of an individual interpersonal contact number prediction system according to an embodiment of this application; Figure 3 This is a schematic diagram of the device structure according to an embodiment of this application; Figure 4 This is a schematic diagram of the structure of the storage medium according to an embodiment of this application. Detailed Implementation

[0017] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0018] The terms "first," "second," and "third" in this application are for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first," "second," or "third" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified. All directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of this application are only used to explain the relative positional relationships and movements between components in a specific orientation (as shown in the figures). If the specific orientation changes, the directional indications also change accordingly. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0019] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0020] Specifically, please refer to Figure 1 This is a flowchart of an individual interpersonal contact number prediction method according to an embodiment of this application. The individual interpersonal contact number prediction method according to an embodiment of this application includes the following steps: S100: Obtain questionnaire survey data on individual residents' travel activities and interpersonal contacts, and perform preprocessing such as anomaly detection and invalid data removal on the questionnaire survey data; In this step, the questionnaire survey data includes, but is not limited to, residents' basic statistical information, travel activity data such as frequency and location of trips, and relevant interpersonal contact information such as the number of people contacted, the locations of contact, and the types of contact. Because the questionnaire survey data may contain abnormal or invalid data, preprocessing such as anomaly detection and invalid data removal is required to avoid overfitting of the model or a decrease in prediction accuracy due to abnormal or invalid data. Specific processing methods include: S101: Review the questionnaire data, screen out invalid questionnaire data that has logical contradictions, highly consistent options, or obvious regularities in the answers, and remove invalid questionnaire data; in order to ensure the objectivity and scientific nature of the data analysis process, the data review process can be carried out under the guidance of statistical experts.

[0021] S102: Outlier detection is performed on questionnaire survey data based on the Raida criterion, and data containing outliers is removed. The Raida criterion applies to data that follows or approximately follows a normal distribution, using three standard deviations as the threshold for outlier detection; data exceeding this threshold are considered outliers. The specific rules for the Raida criterion are as follows: if the residuals of the questionnaire survey data satisfy… Then determine These are outliers. Because the original questionnaire data distribution may deviate from normality, the questionnaire data is logarithmically transformed to approximate a normal distribution before applying the Laida criterion.

[0022] S110: Extract individual resident attributes and travel activity characteristics from preprocessed questionnaire data; In this step, individual attribute characteristics can be directly obtained from the preprocessed questionnaire data. These characteristics include, but are not limited to, information such as the individual resident's gender, age, and monthly income, as shown in Table 1 below. Table 1 Individual Attribute Characteristics

[0023] In Table 1, √ represents features that can be directly extracted from the questionnaire survey data.

[0024] Travel activity characteristics can be calculated based on an individual's travel locations on a given day, reflecting the individual resident's travel frequency, activity range, and movement patterns, providing data support for further modeling. Travel activity characteristics include, but are not limited to, turning radius, furthest distance from home, movement entropy, frequency of activity type on the day, student / employment status, attendance status, commuting distance, frequency of other activities on the day, and status of other activities on the day, as detailed in Table 2 below: Table 2 Characteristics of Individual Travel Activities

[0025] In this context, √ indicates that features can be directly extracted from the questionnaire data.

[0026] Specifically, the definitions of each travel activity characteristic include: The radius of gyration is used to measure an individual's range of activity, and the formula for calculation is: (1) in, Radius of gyration The total number of trips For the first The displacement distance of a trip (usually referring to the straight-line distance from the starting point to the destination).

[0027] The furthest distance from home refers to the maximum geographical distance between an individual's home address and all travel destinations (or trajectory points) within a day. The calculation formula is: (2) in: It represents the total number of trips an individual makes within a specified time period.

[0028] Mobility entropy is an indicator that measures the complexity and diversity of an individual's travel behavior. The formula for its calculation is: (3) in: For the moving entropy, For the number of travel destinations, For individual access Duration of time to each destination.

[0029] Daily activity frequency refers to the total number of all types of activities that an individual engages in on a given day. The "Student / Employed" status identifier is used to indicate whether an individual is currently a student or an employee (value: 0 - No, 1 - Yes). The attendance status indicator for the day is used to indicate whether an individual has attended school or work on the statistical day (value: 0 - no, 1 - yes). The daily commute distance refers to the distance an individual travels to or from school on the day the statistics are compiled. The frequency of other activities on the same day refers to the total number of other activities an individual engages in on the day of the statistics, excluding going to school or work. The "Other Activities on the Day" status flag is used to indicate whether an individual has any other activities besides going to school or work on the day of the statistics (value: 0 - no, 1 - yes).

[0030] S120: Based on demographic and travel activity characteristics, a machine learning algorithm is applied to construct and train an individual interpersonal contact pattern prediction model. In this step, machine learning algorithms are a typical data-driven approach. Their basic idea is to build an empirical model by learning from a training dataset, thereby enabling the prediction and analysis of new data. Compared to traditional methods, machine learning models have fewer assumptions and constraints regarding data distribution, allowing for flexible integration and processing of various types of big data, thus achieving more accurate predictions. This application's embodiment treats each community resident extracted from questionnaire survey data as a sample and uses machine learning methods to solve the problem of predicting individual interpersonal contact patterns. First, the questionnaire survey data is divided into training and testing sets, and the extracted individual attribute features and travel activity features are used as input variables to train the machine learning model. Based on this, an individual interpersonal contact pattern prediction model is constructed to predict the number of contacts an individual has with people of different age groups in different scenarios.

[0031] This application uses machine learning models to predict individual interpersonal contact patterns. These machine learning models include, but are not limited to, Lasso (Least Absolute Shrinkage and Selection Operator) regression models, ElasticNet regression models, CatBoost (Categorical Boosting) models, LightGBM (Light Gradient Boosting Machine) models, XGBoost (eXtreme Gradient Boosting) models, RandomForest, and Multilayer Perceptron (MLP). Among these, the Lasso regression model is a linear model that adds an L1 norm regularization term to the loss function of standard linear regression. This regularization term can compress the coefficients of certain features to zero, thereby achieving feature selection and producing a sparse model. The ElasticNet regression model is a linear model that combines Lasso regression (L1 regularization) and Ridge regression (L2 regularization). By simultaneously introducing two regularization terms, it can maintain stability when there is a high correlation between features and simultaneously perform feature selection. CatBoost is a machine learning algorithm based on gradient boosting decision trees. It can efficiently and directly process categorical features without extensive preprocessing. By using key techniques such as ordered boosting and objective statistics, it effectively avoids gradient bias and prediction offset, performing excellently on datasets with a large number of categorical features. LightGBM is an efficient gradient boosting framework that employs a histogram-based decision tree algorithm and introduces gradient one-sided sampling (GOSS) and mutually exclusive feature binding (EFB) techniques. This significantly improves training speed and reduces memory consumption while maintaining accuracy, making it particularly suitable for large-scale datasets. XGBoost is a scalable gradient boosting decision tree algorithm that controls model complexity by introducing a regularization term into the objective function, effectively preventing overfitting. XGBoost performs a second-order Taylor expansion of the objective function and supports feature parallelism and data parallelism, resulting in excellent performance in both accuracy and computational efficiency. Random forest is an ensemble learning method that enhances the model's prediction accuracy by constructing multiple decision trees and combining their prediction results. Each tree is trained based on a random subset of samples and a random subset of features, exhibiting strong resistance to overfitting and achieving good results in both classification and regression tasks.MLP is a basic feedforward artificial neural network that consists of an input layer, an output layer, and at least one hidden layer. Each neuron performs a linear transformation on the output of the previous layer and outputs the result through a non-linear activation function (such as ReLU or Sigmoid), enabling the network to learn and simulate complex non-linear relationships.

[0032] Furthermore, during model training, this embodiment performs model parameter tuning in two stages: first, the default parameters of each machine learning model are used as a baseline; then, the Randomized SearchCV (RSCCV) method is employed for hyperparameter optimization to improve model performance. RSCCV is an efficient hyperparameter optimization algorithm that, by randomly sampling and evaluating a fixed number of parameter combinations from a specified parameter distribution, can discover high-performance parameter combinations with a higher probability and in a shorter time, making it particularly suitable for high-dimensional parameter spaces. Its basic optimization process includes: Parameter space: Specifies a probability distribution for each hyperparameter, such as a uniform distribution or a log-uniform distribution; Random sampling: randomly selecting from this parameter space Group hyperparameter configuration, among which The number of iterations is preset. Evaluation and Selection: Evaluate the model performance for each hyperparameter configuration on the validation set, and finally select the hyperparameter configuration with the best performance, which can be expressed as: (4) in: This indicates the optimal hyperparameter configuration. This represents a set of hyperparameter configurations. It is a predefined joint probability distribution of hyperparameters. Is it using hyperparameter configuration? The performance evaluation function of the model on the validation set. It is the total number of random samples. This indicates the search for a performance evaluation function. Maximize hyperparameter configuration .

[0033] Furthermore, this application uses the coefficient of determination (R²), mean absolute percentage error (MAPE), mean absolute error (MAE), and root mean square error (RMSE) as evaluation indicators for the predictive performance and error of the individual interpersonal contact pattern prediction model. Among them, This indicates the proportion of data variation explained by the model, ranging from 0 to 1. The closer to 1, the better the model fits the data. The calculation formula is as follows: (5) in, and These are the actual value and the predicted value, respectively. It is the average of the actual values. This represents the number of samples.

[0034] RMSE measures the error between model predictions and actual values. A smaller RMSE value indicates a smaller prediction error and higher model accuracy. The formula for calculating RMSE is as follows: (6) in, and These are the actual value and the predicted value, respectively. This represents the number of samples.

[0035] MAPE (Modular Average Expectation) reflects the average relative deviation between predicted and actual values ​​as a percentage; a lower value indicates higher prediction accuracy. The MAPE calculation formula is: (7) MAE represents the average absolute difference between predicted and actual values. It is insensitive to outliers; a smaller value indicates better model error control. The MAE calculation formula is: (8) It should be noted that the machine learning algorithm used in the embodiments of this application can also be changed from the random forest algorithm to other machine learning or deep learning algorithms such as support vector regression, feedforward neural network or elastic network.

[0036] S130: Use the trained individual interpersonal contact pattern prediction model to predict the number of people an individual will have contact with, and output the predicted value of the number of people an individual will have contact with.

[0037] To verify the effectiveness of the model, this application uses Shenzhen, China as the research area and employs survey data on Shenzhen residents' travel activities and interpersonal contact as the research data source to conduct experiments on the interpersonal contact pattern prediction model. The experimental results are shown in Table 3 below: Table 3 Comparison of Experimental Results

[0038] Table 3 shows that the performance of the seven models used in this application was comprehensively evaluated and compared. Overall, all models showed certain predictive capabilities, but performance varied. Ensemble learning methods generally outperformed linear models and multilayer perceptrons. In terms of the ability to explain variance, Random Forest and XGBoost performed best, followed by CatBoost and LightGBM, while linear models such as Lasso and ElasticNet, as well as MLP, were relatively weak in goodness of fit. Regarding error control, Random Forest outperformed other models in both MAE and RMSE, and XGBoost also performed exceptionally well in MAE. Although MLP had the lowest MAPE, its RMSE and R² showed that its overall predictive stability was inferior to that of tree models. Overall, ensemble learning methods demonstrated superior fitting ability and error robustness across multiple metrics. Experimental results show that this application can achieve accurate predictions from individual characteristics to contact patterns.

[0039] Please see Figure 2 This is a schematic diagram of the individual interpersonal contact number prediction system according to an embodiment of this application. The individual interpersonal contact number prediction system 40 according to an embodiment of this application includes: Data acquisition module 41: Used to acquire questionnaire survey data on individual residents' travel activities and interpersonal contacts; Feature extraction module 42: used to extract individual attribute features and travel activity features of individual residents from the questionnaire survey data; Model building module 43: used to build and train an individual interpersonal contact pattern prediction model based on the demographic characteristics and travel activity characteristics, using machine learning algorithms; Interpersonal contact prediction module 44: Used to predict the number of interpersonal contacts of an individual using a trained individual interpersonal contact pattern prediction model, and output the predicted value of the number of interpersonal contacts of an individual.

[0040] It should be noted that since the information interaction and execution process between the system embodiments of this application and the above-mentioned methods / devices / modules / units are based on the same concept, their specific functions and technical effects can be found in the method embodiments section, and will not be repeated here.

[0041] Based on the above, the individual interpersonal contact number prediction method and system of this application extracts individual attribute features and travel activity features from questionnaire survey data, then uses machine learning algorithms to construct an individual interpersonal contact pattern prediction model, and finally outputs the number of individual interpersonal contacts, achieving accurate prediction from individual characteristics to contact patterns. This application, by integrating dynamic travel activity features with static demographic attributes and utilizing machine learning modeling, can capture refined individual behavior, achieving a more reliable prediction of the number of individual contacts. This not only provides high-precision data support for infectious disease transmission models, improving the accuracy of disease transmission risk assessment and enhancing the pertinence and effectiveness of epidemic prevention and control strategies, but also provides a transferable practical example for similar studies in other cities or regions.

[0042] Please see Figure 3 This is a schematic diagram of the device structure according to an embodiment of this application. The device 50 includes: Memory 51 storing executable program instructions; Processor 52 connected to memory 51; The processor 52 is used to call the executable program instructions stored in the memory 51 and perform the following steps: obtain questionnaire survey data on individual residents' travel activities and interpersonal contact; extract individual attribute characteristics and travel activity characteristics of individual residents from the questionnaire survey data; based on the demographic characteristics and travel activity characteristics, apply machine learning algorithms to construct and train an individual interpersonal contact pattern prediction model; use the trained individual interpersonal contact pattern prediction model to predict the number of individual interpersonal contacts, and output the predicted value of the number of individual interpersonal contacts.

[0043] The processor 52 can also be referred to as a CPU (Central Processing Unit). The processor 52 may be an integrated circuit chip with signal processing capabilities. The processor 52 can also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), an off-the-shelf programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. A general-purpose processor can be a microprocessor or any conventional processor.

[0044] Please see Figure 4This is a schematic diagram of the structure of the storage medium in an embodiment of this application. The storage medium in this embodiment stores program instructions 61 capable of implementing the following steps: acquiring questionnaire survey data on individual residents' travel activities and interpersonal contacts; extracting individual attribute characteristics and travel activity characteristics of individual residents from the questionnaire survey data; constructing and training an individual interpersonal contact pattern prediction model based on the demographic characteristics and travel activity characteristics using machine learning algorithms; predicting the number of individual interpersonal contacts using the trained individual interpersonal contact pattern prediction model, and outputting the predicted value of the number of individual interpersonal contacts. The program instructions 61 can be stored in the aforementioned storage medium in the form of a software product, including several instructions to cause a device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods in various embodiments of this application. The aforementioned storage medium includes various media capable of storing program instructions, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks, or terminal devices such as computers, servers, mobile phones, and tablets. The server can be a standalone server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms.

[0045] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interfaces, apparatuses, or units, and may be electrical, mechanical, or other forms.

[0046] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units described above can be implemented in hardware or as software functional units. The above are merely embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made based on the description and drawings of this application, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A method for predicting the number of interpersonal contacts an individual may have, characterized in that, include: Obtain questionnaire survey data on individual residents' travel activities and interpersonal contacts; Individual resident attributes and travel activity characteristics were extracted from the questionnaire survey data. Based on the aforementioned demographic and travel activity characteristics, a machine learning algorithm is applied to construct and train an individual interpersonal contact pattern prediction model. The trained individual interpersonal contact pattern prediction model is used to predict the number of interpersonal contacts for an individual, and the predicted value of the number of interpersonal contacts for an individual is output.

2. The method for predicting the number of interpersonal contacts of an individual according to claim 1, characterized in that, After obtaining questionnaire survey data on individual residents' travel activities and interpersonal contacts, the process also includes: The questionnaire survey data is preprocessed, including anomaly detection and invalid data removal.

3. The method for predicting the number of interpersonal contacts of an individual according to claim 2, characterized in that, The preprocessing of the questionnaire data includes: The questionnaire data is reviewed, and invalid questionnaire data that contains logical inconsistencies, highly consistent options, or obvious patterns of responses is filtered out and removed. Outlier detection was performed on the questionnaire data based on the Raida criterion, and questionnaire data containing outliers were removed.

4. The method for predicting the number of interpersonal contacts of an individual according to any one of claims 1 to 3, characterized in that, The individual attribute characteristics include the individual resident's gender, age, and monthly income; the travel activity characteristics include turning radius, furthest distance from home, mobility entropy, frequency of daily activity types, student / employment status, attendance status, commuting distance, frequency of other daily activities, and status of other daily activities; wherein, the turning radius is used to measure the individual's activity range, the furthest distance from home refers to the maximum geographical distance between the individual's home address and all travel destinations or trajectory points within a day, the mobility entropy is an indicator of the complexity and diversity of the individual's travel behavior, and the frequency of daily activities refers to the individual's... The total number of all types of activities that occurred on the day is counted. The "Study / Employment Status" identifier is used to identify whether an individual is currently a student or an employee. The "Daily Attendance Status" identifier is used to identify whether an individual attended school or work on the day of the count. The "Daily Commuting Distance" refers to the commuting distance an individual traveled to school or work on the day of the count. The "Daily Other Activity Frequency" refers to the total number of other activities an individual engaged in on the day of the count, excluding going to school or work. The "Daily Other Activity Status" identifier is used to identify whether an individual engaged in other activities besides going to school or work on the day of the count.

5. The method for predicting the number of interpersonal contacts of an individual according to claim 4, characterized in that, The process of constructing and training an individual interpersonal contact pattern prediction model based on the demographic and travel activity characteristics using machine learning algorithms is as follows: The questionnaire data is divided into training and testing sets, and the individual attribute features and travel activity features are used as input variables to train a machine learning model. The machine learning model includes Lasso regression model, ElasticNet regression model, CatBoost model, LightGBM model, XGBoost model, random forest, and multilayer perceptron.

6. The method for predicting the number of interpersonal contacts of an individual according to claim 5, characterized in that, The method of constructing and training an individual interpersonal contact pattern prediction model based on the demographic and travel activity characteristics using machine learning algorithms also includes: During model training, the default parameters of each machine learning model are first used as a baseline, followed by hyperparameter optimization using a random cross-reference method. The optimization process of the random cross-reference method includes: Parameter space: Specifies a probability distribution for each hyperparameter; Random sampling: randomly selecting from the parameter space Group hyperparameter configuration, among which The number of iterations is preset. Evaluation and selection: Evaluate the model performance of each hyperparameter configuration on the validation set and select the hyperparameter configuration with the best performance.

7. The method for predicting the number of interpersonal contacts of an individual according to claim 6, characterized in that, The method of constructing and training an individual interpersonal contact pattern prediction model based on the demographic and travel activity characteristics using machine learning algorithms also includes: The coefficient of determination, mean absolute percentage error, mean absolute error, and root mean square error are used as evaluation indicators for the predictive performance and error of the individual interpersonal contact pattern prediction model. The coefficient of determination represents the proportion of data variation explained by the model, the mean absolute percentage error is used to measure the error between the model's predicted value and the actual value, the mean absolute error reflects the average relative deviation between the predicted value and the actual value in the form of a percentage, and the root mean square error represents the average value of the absolute difference between the predicted value and the actual value.

8. A system for predicting the number of interpersonal contacts for an individual, characterized in that, include: Data acquisition module: used to acquire questionnaire survey data on individual residents' travel activities and interpersonal contacts; Feature extraction module: used to extract individual attribute features and travel activity features of individual residents from the questionnaire survey data; Model building module: used to build and train an individual interpersonal contact pattern prediction model based on the demographic characteristics and travel activity characteristics, using machine learning algorithms; Interpersonal contact prediction module: This module is used to predict the number of interpersonal contacts an individual will have using a pre-trained individual interpersonal contact pattern prediction model, and outputs the predicted number of interpersonal contacts.

9. A device, characterized in that, The device includes a processor and a memory coupled to the processor, wherein, The memory stores program instructions for implementing the method for predicting the number of interpersonal contacts of an individual; The processor is used to execute the program instructions stored in the memory to control the method for predicting the number of interpersonal contacts of an individual.

10. A storage medium, characterized in that, The system stores processor-executable program instructions for performing the individual interpersonal contact number prediction method.

Images