CN114464329A - Urban epidemic situation space-time prediction method, system, terminal and storage medium - Google Patents
Urban epidemic situation space-time prediction method, system, terminal and storage medium Download PDFInfo
- Publication number
- CN114464329A CN114464329A CN202111675147.7A CN202111675147A CN114464329A CN 114464329 A CN114464329 A CN 114464329A CN 202111675147 A CN202111675147 A CN 202111675147A CN 114464329 A CN114464329 A CN 114464329A
- Authority
- CN
- China
- Prior art keywords
- infectious disease
- time
- data
- regions
- prediction
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 68
- 238000003860 storage Methods 0.000 title claims abstract description 12
- 208000035473 Communicable disease Diseases 0.000 claims abstract description 145
- 208000015181 infectious disease Diseases 0.000 claims abstract description 139
- 230000015654 memory Effects 0.000 claims abstract description 27
- 238000013528 artificial neural network Methods 0.000 claims abstract description 26
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 15
- 238000012545 processing Methods 0.000 claims abstract description 12
- 239000011159 matrix material Substances 0.000 claims description 32
- 238000004364 calculation method Methods 0.000 claims description 15
- 230000006870 function Effects 0.000 claims description 7
- 230000006403 short-term memory Effects 0.000 claims description 6
- 230000004913 activation Effects 0.000 claims description 5
- 230000002776 aggregation Effects 0.000 claims description 5
- 238000004220 aggregation Methods 0.000 claims description 5
- 230000007787 long-term memory Effects 0.000 claims description 5
- 238000010276 construction Methods 0.000 claims description 3
- 238000013480 data collection Methods 0.000 claims description 3
- 238000007405 data analysis Methods 0.000 claims description 2
- 230000004927 fusion Effects 0.000 claims description 2
- 230000008447 perception Effects 0.000 abstract description 4
- 238000011161 development Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 238000010801 machine learning Methods 0.000 description 6
- 201000010099 disease Diseases 0.000 description 5
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 5
- 230000003993 interaction Effects 0.000 description 5
- 238000013179 statistical model Methods 0.000 description 5
- 230000008859 change Effects 0.000 description 4
- 230000002265 prevention Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 230000005180 public health Effects 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 238000013135 deep learning Methods 0.000 description 3
- 230000001932 seasonal effect Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 230000036541 health Effects 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000010946 mechanistic model Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 239000013598 vector Substances 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 238000011217 control strategy Methods 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 230000005541 medical transmission Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000007500 overflow downdraw method Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000001902 propagating effect Effects 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 230000011664 signaling Effects 0.000 description 1
- 230000007480 spreading Effects 0.000 description 1
- 238000003892 spreading Methods 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/80—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for detecting, monitoring or modelling epidemics or pandemics, e.g. flu
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Biomedical Technology (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Business, Economics & Management (AREA)
- Public Health (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Economics (AREA)
- Medical Informatics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Development Economics (AREA)
- Databases & Information Systems (AREA)
- Game Theory and Decision Science (AREA)
- Pathology (AREA)
- Epidemiology (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application relates to a city epidemic situation space-time prediction method, a system, a terminal and a storage medium. The method comprises the following steps: collecting individual movement track data and infectious disease case data in a city; processing the individual movement track data, extracting population movement flow among all areas, dividing the population movement flow according to time attributes, and acquiring population movement relations under all time attributes based on the adjacent relations of the areas; calculating the similarity of the infectious disease case data among the regions by using a histogram-based similarity algorithm, and acquiring the position attention relationship of the regions according to the similarity of the infectious disease case data; and constructing an infectious disease space-time prediction model based on the graph neural network and the long-short term memory network according to population movement relations, proximity relations and position attention relations. The method and the device can model the time dependence and the space dependence in the data of the infectious disease cases, and improve the space perception capability and the prediction performance of the space-time prediction model.
Description
Technical Field
The application belongs to the technical field of infectious disease calculation, and particularly relates to a city epidemic situation space-time prediction method, a system, a terminal and a storage medium.
Background
Due to the increased availability of public health monitoring data and the development of sophisticated methods, infectious disease prediction has become increasingly prominent in recent years, and real-time prediction of disease infection remains plagued by the lack of current estimates of mobility and interaction patterns. While fluidity and interaction patterns are key drivers of disease transmission. The prediction of the infectious disease incidence trend is an important problem in the field of public health, and key information of infectious diseases, such as peak intensity and outbreak time, can be found, tracked and predicted in time and is important for effectively making prevention and control strategies and implementing intervention measures. According to the modeling principle and purpose, the infectious disease model can be divided into two categories, namely a mechanism model and a non-mechanism model.
Taking seasonal infectious diseases as an example, the seasonal infectious diseases are monitored and predicted, and corresponding prevention and control preparations are made in time, so that the seasonal infectious diseases and infectious disease pandemic control are very important. The existing infectious disease prediction modeling method comprises a Gaussian process model, an LSTM (Long Short-Term Memory) neural network model and the like. However, the existing infectious disease prediction modeling methods include modeling based on a traditional statistical model and modeling based on traditional machine learning, and the infectious disease prediction methods based on the traditional statistical model require data to meet strict assumptions, such as the assumption of stationarity of time series. However, in the real world, these assumptions are not easily satisfied. Most infectious disease prediction methods based on traditional machine learning still rely on feature engineering, and accurate prediction cannot be achieved on difficult prediction problems. In addition, other characteristics from space dimensionality are generally ignored in the method, and further the space-time dependency contained in the space-time data cannot be modeled, so that the model prediction performance is limited.
The main core of the mechanistic model is the propagation dynamics model, the most commonly used being the bin model. The compartment model divides the population or other hosts in the same unit into different compartments corresponding to the population or other hosts according to the infection state of an infected person, and then simulates the disease development dynamics of the different compartments based on the disease propagation characteristics. Typical representatives of the above are susceptible-infected-recovered (SIR) and susceptible-latent-infected-recovered (SEIR) models. Such models are often used to predict long-term trends in disease, and to deduce the effects of different interventions. However, the modeling is complex, the updating is slow, the uncertainty of the model parameters can greatly influence the accuracy of the model, and the real-time accurate prediction of the short-term development trend of the infectious disease is difficult to realize.
Non-mechanistic models include statistical models and machine learning models, and conventional statistical models require that data satisfy strict assumptions (e.g., a time series stationarity assumption). However, in the real world, these assumptions are not easily satisfied. Although infectious disease prediction methods based on traditional machine learning are data-driven and do not need to meet strict assumptions about data, most traditional machine learning methods still rely on feature engineering and cannot achieve accurate prediction on difficult prediction problems. In addition, in the modeling process, regardless of the infectious disease prediction method based on the traditional statistical model or the infectious disease prediction method based on the traditional machine learning model, other features from the space dimension are generally ignored, and further the space-time dependency contained in the space-time data cannot be modeled. This modeling approach limits the predictive performance of these methods. The infectious disease prediction method based on deep learning can automatically learn nonlinear characteristics from the spatio-temporal data, and has greater performance advantages. Some studies have designed a prediction framework based on deep learning from multiple perspectives to better model the spatio-temporal dependencies implied in spatio-temporal data to improve the prediction performance of the model. However, these studies mostly do not consider spatial interaction caused by population movement and a spatio-temporal spreading pattern of contagion inside cities when modeling spatial dependency between different regions, and such methods have difficulty in accurately predicting the epidemic tendency of contagion inside cities due to the lack of information.
Disclosure of Invention
The application provides a method, a system, a terminal and a storage medium for urban epidemic situation space-time prediction, which aim to solve at least one of the technical problems in the prior art to a certain extent.
In order to solve the above problems, the present application provides the following technical solutions:
a city epidemic situation space-time prediction method comprises the following steps:
collecting individual movement track data and infectious disease case data in a city;
processing the individual movement track data through a data driving method, extracting population movement flow among all areas, dividing the population movement flow according to time attributes, and acquiring population movement relations under all time attributes based on the adjacent relations of the areas;
calculating the similarity of the infectious disease case data among the regions by using a histogram-based similarity algorithm, and acquiring the position attention relationship of the regions according to the similarity of the infectious disease case data; constructing an infectious disease space-time prediction model based on a graph neural network and a long-short term memory network according to the population movement relationship, the proximity relationship and the position attention relationship;
and inputting the urban infectious disease case data into the infectious disease space-time prediction model, and obtaining the urban infectious disease prediction result through the infectious disease space-time prediction model.
The technical scheme adopted by the embodiment of the application further comprises the following steps: the population movement traffic is divided according to the time attributes, and the population movement relationship under each time attribute is obtained based on the proximity relationship of the region specifically:
the time attribute comprises a weekday, a weekend, or a holiday; respectively converting the population movement traffic of each time attribute into a weighted directed graph G (V, E), wherein V represents a node set, E represents an edge set, a vertex represents each area in the city, and an edge is used for capturing a movement mode; and judging whether a connecting edge exists between two nodes in the directed graph or not according to whether the two regions are in mutual contact or not based on the proximity relation of the regions, so as to obtain a population movement relation adjacency matrix corresponding to the proximity relation.
The technical scheme adopted by the embodiment of the application further comprises the following steps: the calculating the similarity of the infectious disease case data among the regions by using a histogram-based similarity algorithm, and the obtaining the position attention relationship of the regions according to the similarity of the infectious disease case data comprises:
constructing corresponding histograms for the infectious disease case data of each region respectively;
and calculating the similarity of the infectious disease case data between the two regions, if the similarity of the infectious disease case data between the two regions is higher than a set threshold value, considering that the infectious disease outbreak trends of the two regions are similar and have correlation, constructing a connecting edge on a graph between the two regions in the directed graph, and generating a adjacency matrix representing the position attention relationship between all the regions.
The technical scheme adopted by the embodiment of the application further comprises the following steps: the calculation formula of the position attention relationship is as follows:
wherein theta is a threshold value for establishing a connecting edge based on a similarity algorithm of the histogram; when the similarity w between the region i and the region ji,jAnd when the distance is higher than the threshold value theta, creating a connecting edge between the area i and the area j, and finally obtaining an adjacent matrix of the attention relation of the corresponding positions of each area in the city.
The technical scheme adopted by the embodiment of the application further comprises the following steps: the building of the infectious disease space-time prediction model based on the graph neural network and the long-short term memory network according to the population movement relationship, the proximity relationship and the position attention relationship comprises the following steps:
inputting the graph structure into a graph neural network, wherein the graph neural network normalizes the directed graph by using a neighborhood aggregation method, and the incoming edge weighting of each node is equal to 1:
wherein HiIs a matrix containing the node representations of the previous layer, the initial H0Set as history data representing case changes of infectious diseases in each region, WiA trainable parameter matrix representing the i-th layer, f is non-linearThe function is activated.
The technical scheme adopted by the embodiment of the application further comprises the following steps: the building of the infectious disease space-time prediction model based on the graph neural network and the long-short term memory network according to the population movement relationship, the proximity relationship and the position attention relationship further comprises the following steps:
obtaining a representation sequence h by using a message propagation neural network at each time stepi,t-n,hi,t-n+1,...,hi,t-1Will represent the sequence hi,t-n,hi,t-n+1,...,hi,t-1Inputting the long-term and short-term memory network, and extracting the time sequence relation in the network; the long-short term memory network calculation formula is as follows:
Xi,t=LSTM(hi,t-n,hi,t-n+1,...,hi,t-1) Wherein, Xi,tRepresenting predicted case data of infectious disease in the ith region in the t-th time period, hi,t-1Representing the infectious disease case data of the ith area in the t-1 th time period.
The technical scheme adopted by the embodiment of the application further comprises the following steps: the method comprises the following steps of inputting urban infectious disease case data into the infectious disease space-time prediction model, and obtaining urban infectious disease prediction results through the infectious disease space-time prediction model:
fusing a proximity relation, a population movement relation and a position attention relation through an infectious disease space-time prediction model to obtain a city infectious disease prediction result; the fusion mode specifically comprises the following steps:
wherein, Wadj、WodAnd WatIs a matrix of parameters that needs to be trained,andinfectious disease prediction node at time t obtained based on adjacency matrix, population movement flow matrix and position attention matrixAs a result, tanh is the activation function,and (4) obtaining the final prediction result of the whole time-space prediction model at the time t.
Another technical scheme adopted by the embodiment of the application is as follows: a city epidemic situation space-time prediction system comprises:
a data collection module: the system is used for collecting individual movement track data and infectious disease case data in a city;
a flow calculation module: the system comprises a data driving method, a data acquisition method and a data analysis method, wherein the data driving method is used for processing the individual movement track data, extracting population movement flow among all areas, dividing the population movement flow according to time attributes, and acquiring population movement relations under all the time attributes based on the proximity relations of the areas;
a similarity calculation module: the method comprises the steps of calculating the similarity of infectious disease case data among all regions by using a histogram-based similarity algorithm, and acquiring the position attention relationship of all regions according to the similarity of the infectious disease case data;
a model construction module: the system is used for constructing an infectious disease space-time prediction model based on a graph neural network and a long-short term memory network according to the population movement relationship, the proximity relationship and the position attention relationship;
an infectious disease prediction module: and the system is used for inputting the urban infectious disease case data into the infectious disease space-time prediction model and obtaining the urban infectious disease prediction result through the infectious disease space-time prediction model.
The embodiment of the application adopts another technical scheme that: a terminal comprising a processor, a memory coupled to the processor, wherein,
the memory stores program instructions for implementing the urban epidemic situation spatiotemporal prediction method;
the processor is configured to execute the program instructions stored in the memory to control urban epidemic spatiotemporal prediction.
The embodiment of the application adopts another technical scheme that: a storage medium storing program instructions executable by a processor to perform the method of urban epidemic spatiotemporal prediction.
Compared with the prior art, the embodiment of the application has the advantages that: according to the urban epidemic situation space-time prediction method, the urban epidemic situation space-time prediction system, the urban population movement flow is extracted through the urban individual movement trajectory data, the population movement flow in a certain time period is divided according to the time attributes, the proximity relation, the population movement relation and the position attention relation among all the areas are obtained according to the population movement flow with different time attributes, an infectious disease space-time prediction model based on a graph neural network and a long-short term memory network is built according to the proximity relation, the population movement relation and the position attention relation, and the infectious disease trend is finely predicted through the infectious disease space-time prediction model. The embodiment of the application can accurately capture the influence of population interaction and movement under different time attributes on the spread and transmission of the infectious diseases in the city by dividing the time attributes of population movement flow, fully utilize local spatial context information of each region, consider various spatial relationships among the regions, therefore, time dependency and space dependency in infectious disease case data are better modeled, the space perception capability and the model prediction performance of the infectious disease space-time prediction model are greatly improved, the prediction of the urban infectious disease development situation with higher space resolution is realized, the fine analysis of the infectious disease epidemic situation is completed, the government and the public health department are helped to timely and accurately monitor the urban infectious disease development situation, the epidemic situation prevention and control intervention is performed in a targeted manner, and the life health safety of people can be guaranteed to the maximum extent.
Drawings
FIG. 1 is a flow chart of a city epidemic situation spatio-temporal prediction method according to an embodiment of the present application;
FIG. 2 is a schematic structural diagram of a city epidemic situation spatiotemporal prediction system according to an embodiment of the present application;
fig. 3 is a schematic structural diagram of a terminal according to an embodiment of the present application;
fig. 4 is a schematic structural diagram of a storage medium according to an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
Aiming at the defects of the prior art, the urban epidemic situation space-time prediction method disclosed by the embodiment of the application constructs a space-time prediction model based on a graph neural network and a long-short term memory network based on the proximity relation, population movement relation and position attention relation among all regions in a city so as to better model the time dependence and space dependence in space-time data, and considers various space relations among the regions and constructs a corresponding relational graph structure when modeling the space dependence; when the time dependency is modeled, the local spatial context information of each region is fully utilized to improve the prediction performance of the model; and dividing population movement flow into different time attributes according to population movement modes of infectious disease susceptible populations, and respectively constructing directed graph structures with different time attributes so as to improve the space-time perception capability of the space-time prediction model and further improve the prediction performance of the model.
Specifically, please refer to fig. 1, which is a flowchart of the urban epidemic situation spatio-temporal prediction method according to the embodiment of the present application. The urban epidemic situation space-time prediction method comprises the following steps:
s1: acquiring urban infectious disease case data, and collecting urban individual movement track data by using mobile equipment;
in this step, the infectious disease case data includes, but is not limited to, data such as the number of infected cases, the number of dead cases, clinical symptoms, and information on hospital staff. The individual movement trace data includes, but is not limited to, a mobile phone number, a signaling time stamp, mobile phone position data and the like of each individual. In order to better predict the urban internal infectious disease incidence trend of susceptible people such as minors, large-scale individual movement trajectory data infectious disease cases are obtained in the embodiment of the application. Mobile devices include, but are not limited to, devices such as cell phones or smart watches.
S2: processing the individual movement track data through a data driving method, extracting population movement flow among different areas, and acquiring a city population movement relation;
in this step, the processing of the individual movement trajectory data by the data driving method specifically includes: and (3) extracting population movement flow among different areas through large-scale mobile phone position data, and capturing the urban population movement relationship.
S3: dividing population movement traffic according to the time attributes, constructing a directed graph of the population movement traffic with different time attributes, judging whether a connecting edge exists between two nodes in the directed graph or not based on the proximity relation of areas, and generating an adjacency matrix of the population movement relation under different time attributes;
in this step, the time attribute includes, but is not limited to, weekday, weekend, or holiday. Assuming a city is given, a graph structure with time interval units of day, week or month is created, population movement traffic is divided into population movement traffic with two time attributes according to working day and weekend, the population movement traffic with each time attribute is converted into a weighted directed graph G ═ V, E, wherein V represents a node set, E represents an edge set, vertexes represent areas in the city, and edges are used for capturing movement patterns. For example, the weight from vertex v to vertex uRepresenting the total number of people moving from the area v to the area u on the weekday of the t week, thereby respectively obtaining an adjacency matrix of population movement relations of the city under the two time attributes of the weekday and the weekend of the t weekAnd
further, according to the first law of geography, all things are related to other things, but objects with close distances are compared with distancesObjects far away are more relevant, so the embodiment of the present application considers the spatial relationship between objects close to each other, i.e. the proximity relationship. Based on the proximity relation of the areas, judging whether a connecting edge exists between the two areas (nodes) according to whether the two areas are in mutual contact or not, and obtaining a population movement relation adjacency matrix A corresponding to the proximity relationadj. Specifically, assuming that the regions v and u are adjacent regions, the weight from the vertex v to the vertex u is weightedThe total number of individuals moving to the area u from the area v on the weekday is shown, and the directed graph G can contain the population flowing behavior of each adjacent area in the city. The fluidity of the areas v and u on the working day forms a side, and the side is multiplied by the number of cases in the area v in the period of time to obtain a relative score which indicates how many infected persons may flow from the area v to the area u. It will be understood that the case change pattern between two adjacent regions may also have no fluidity, for example, if regions i and j are adjacent regions, the flow rate of region i is fixed and constantly zero, and the flow rate of region j is dynamic and non-zero, then the case change patterns of the two adjacent regions have no correlation.
S4: constructing a histogram of the infectious disease case data, calculating the similarity of the infectious disease case data among all the regions by using a similarity algorithm based on the histogram, and acquiring an adjacency matrix representing the position attention relationship among all the regions according to the similarity of the infectious disease case data;
in this step, the correlation between two areas may be affected by the geographic distance, i.e., adjacent areas may have similar topographic or climatic features, giving them a similar tendency to catch an infection. However, due to population mobility and similar geographic features, non-adjacent areas may also have potential dependencies, but it is difficult to model all relevant factors of an infection outbreak. Therefore, the present invention calculates the correlation of the infectious disease case data between the respective regions using the histogram-based similarity algorithm. If the similarity of the infectious disease case data of the two regions is high, and accordingly the case change patterns of the two regions are similar, the embodiment of the application captures the case change trend in a more global spatial view by calculating the similarity of the infectious disease case data of the regions which are not adjacent but have strong correlation.
The similarity calculation method based on the histogram specifically comprises the following steps: firstly, constructing corresponding histograms for time series of infectious disease case data samples in different regions respectively; and secondly, calculating the similarity of the infectious disease case data of the two regions, if the similarity of the infectious disease case data of the two regions is higher than a set threshold value, considering that the infectious disease outbreak trends of the two regions are relatively similar and the two regions have relatively strong correlation, and constructing a connecting edge on a graph between the two regions in the directed graph so as to generate an adjacency matrix representing the position attention relationship between all the regions. The calculation formula is as follows:
and theta is a threshold value for establishing a connecting edge based on a similarity algorithm of the histogram. When the similarity w between the region i and the region ji,jWhen the distance is higher than the threshold value theta, a connecting edge is created between the area i and the area j, and finally an adjacent matrix of the attention relation of the corresponding positions of each area in the city is obtained
Specifically, the histogram construction algorithm is as follows:
the histogram-based similarity algorithm is as follows:
similarity calculation method
S5: training an infectious disease spatio-temporal prediction model based on a Graph Neural Network (GNN) and a Long-short-term memory network (LSTM) by utilizing a proximity relation, a population movement relation and a position attention relation;
in this step, the graph neural network is a neighbor aggregation strategy, and the representation vector of a node is calculated by the neighbor node through cyclic aggregation and transfer representation vectors. The framework of the graph Neural Network is a Message Propagating Neural Network (MPNN). The message propagation neural network is a formalized framework of spatial graph convolution. In the embodiment of the application, the graph neural network normalizes the input directed graph structure matrix a by using the following neighborhood aggregation method:
wherein HiIs a matrix containing the node representations of the previous layer, the initial H0Set as history data representing case changes of infectious diseases in each region, WiA trainable parameter matrix representing the i-th layer, f is a non-linear activation function, such as a ReLU function. Normalizing the input directed graph structure matrix A to enable the incoming edge weight of each node to be equal to 1, and obtaining the normalized directed graph structure matrix
The long-short term memory Network is a special Recurrent Neural Network (RNN) that can be used to process sequence data. The long and short term memory network mainly aims to solve the problems of gradient loss and gradient explosion in the long sequence training process. Using an MPNN at each time step to obtain a representation sequence hi,t-n,hi,t-n+1,...,hi,t-1. And inputting the representation sequences into a long-term and short-term memory network, and extracting the time characteristics in the long-term and short-term memory network. The LSTM calculation is expressed as:
Xi,t=LSTM(hi,t-n,hi,t-n+1,...,hi,t-1) (3)
wherein, Xi,tRepresenting predicted case data of infectious disease in the ith region in the t-th time period, hi,t-1And (3) an indication of input characteristics such as infectious disease case data of the ith area in the t-1 th time period.
S6: fusing a proximity relation, a population movement relation and a position attention relation through an infectious disease space-time prediction model to obtain a city infectious disease prediction result;
in order to simultaneously consider the influence of the proximity relation, the population movement relation and the position attention relation on the prediction of the urban internal infectious diseases, the invention adopts a fusion method based on a parameter matrix, which specifically comprises the following steps:
wherein, Wadj、WodAnd WatIs a matrix of parameters that needs to be trained,andrespectively, the infectious disease prediction result at time t obtained based on the adjacency matrix, the population movement traffic matrix and the location attention matrix, tanh is an activation function,the final prediction result of the whole infectious disease space-time prediction model at the moment t is obtained.
According to the method and the device, the infectious disease trend prediction is carried out in an artificial intelligence deep learning mode, and the model parameters can be updated by learning new data after multiple predictions, so that the model is more intelligent and efficient.
Based on the above, the urban epidemic situation space-time prediction method provided by the embodiment of the application extracts urban population movement flow through urban individual movement trajectory data, divides the population movement flow in a certain time period according to time attributes, obtains the proximity relation, population movement relation and position attention relation among various areas according to the population movement flow with different time attributes, constructs an infectious disease space-time prediction model based on a graph neural network and a long-short term memory network according to the proximity relation, the population movement relation and the position attention relation, and makes a refined prediction on infectious disease trend through the infectious disease space-time prediction model. The embodiment of the application can accurately capture the influence of population interaction and movement under different time attributes on the spread and transmission of the infectious diseases in the city by dividing the time attributes of population movement flow, fully utilize local spatial context information of each region, consider various spatial relationships among the regions, therefore, time dependency and space dependency in infectious disease case data are better modeled, the space perception capability and the model prediction performance of the infectious disease space-time prediction model are greatly improved, the prediction of the urban infectious disease development situation with higher space resolution is realized, the fine analysis of the infectious disease epidemic situation is completed, the government and the public health department are helped to timely and accurately monitor the urban infectious disease development situation, the epidemic situation prevention and control intervention is performed in a targeted manner, and the life health safety of people can be guaranteed to the maximum extent.
Please refer to fig. 2, which is a schematic structural diagram of a city epidemic situation spatio-temporal prediction system according to an embodiment of the present application. The urban epidemic situation spatiotemporal prediction system 40 of the embodiment of the application comprises:
the data collection module 41: the system is used for collecting individual movement track data and infectious disease case data in a city;
the flow calculation module 42: the system comprises a data driving method, a data acquisition and processing module and a data processing module, wherein the data driving method is used for processing individual movement track data, extracting population movement flow among all areas, dividing the population movement flow according to time attributes and acquiring population movement relations under all the time attributes based on the proximity relations of the areas;
similarity calculation module 43: the method is used for calculating the similarity of the infectious disease case data among the regions by using a histogram-based similarity algorithm and acquiring the position attention relationship of the regions according to the similarity of the infectious disease case data;
model building module 44: the method is used for constructing an infectious disease space-time prediction model based on a graph neural network and a long-short term memory network according to population movement relations, proximity relations and position attention relations;
infectious disease prediction module 45: the method is used for inputting urban infectious disease case data into an infectious disease space-time prediction model and obtaining urban infectious disease prediction results through the infectious disease space-time prediction model.
Please refer to fig. 3, which is a schematic diagram of a terminal structure according to an embodiment of the present application. The terminal 50 comprises a processor 51, a memory 52 coupled to the processor 51.
The memory 52 stores program instructions for implementing the above-described urban epidemic spatiotemporal prediction method.
The processor 51 is operable to execute program instructions stored in the memory 52 to control the urban epidemic spatiotemporal prediction.
The processor 51 may also be referred to as a CPU (Central Processing Unit). The processor 51 may be an integrated circuit chip having signal processing capabilities. The processor 51 may 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 device, discrete gate or transistor logic, discrete hardware components. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
Please refer to fig. 4, which is a schematic structural diagram of a storage medium according to an embodiment of the present application. The storage medium of the embodiment of the present application stores a program file 61 capable of implementing all the methods described above, where the program file 61 may be stored in the storage medium in the form of a software product, and includes several instructions to enable a computer device (which may be a personal computer, a server, or a network device) or a processor (processor) to execute all or part of the steps of the methods of the embodiments of the present invention. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a mobile hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, or terminal devices, such as a computer, a server, a mobile phone, and a tablet.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. A city epidemic situation space-time prediction method is characterized by comprising the following steps:
collecting individual movement track data and infectious disease case data in a city;
processing the individual movement track data through a data driving method, extracting population movement flow among all areas, dividing the population movement flow according to time attributes, and acquiring population movement relations under all time attributes based on the adjacent relations of the areas;
calculating the similarity of the infectious disease case data among the regions by using a histogram-based similarity algorithm, and acquiring the position attention relationship of the regions according to the similarity of the infectious disease case data;
constructing an infectious disease space-time prediction model based on a graph neural network and a long-short term memory network according to the population movement relationship, the proximity relationship and the position attention relationship;
and inputting the urban infectious disease case data into the infectious disease space-time prediction model, and obtaining the urban infectious disease prediction result through the infectious disease space-time prediction model.
2. The urban epidemic situation spatiotemporal prediction method according to claim 1, wherein the population movement traffic is divided according to time attributes, and the population movement relationships under each time attribute obtained based on the proximity relationship of the area are specifically:
the time attribute comprises a weekday, a weekend, or a holiday; respectively converting the population movement traffic of each time attribute into a weighted directed graph G (V, E), wherein V represents a node set, E represents an edge set, a vertex represents each area in the city, and an edge is used for capturing a movement mode; and judging whether a connecting edge exists between two nodes in the directed graph or not according to whether the two regions are in mutual contact or not based on the proximity relation of the regions, so as to obtain a population movement relation adjacency matrix corresponding to the proximity relation.
3. The urban epidemic situation spatiotemporal prediction method according to claim 2, wherein the calculating the similarity of the infectious disease case data among the regions by using a histogram-based similarity algorithm, and the obtaining the position attention relationship of the regions according to the similarity of the infectious disease case data comprises:
constructing corresponding histograms for the infectious disease case data of each region respectively;
and calculating the similarity of the infectious disease case data between the two regions, if the similarity of the infectious disease case data between the two regions is higher than a set threshold value, considering that the infectious disease outbreak trends of the two regions are similar and have correlation, constructing a connecting edge on a graph between the two regions in the directed graph, and generating a adjacency matrix representing the position attention relationship between all the regions.
4. The urban epidemic situation spatiotemporal prediction method according to claim 3, wherein the calculation formula of the position attention relationship is as follows:
wherein theta is a threshold value for establishing a connecting edge based on a similarity algorithm of the histogram; when the similarity w between the region i and the region ji,jAbove the threshold θ, then in region i and zoneAnd a connecting edge is created between the domains j, and finally an adjacent matrix of the attention relationship of the corresponding positions of each region in the city is obtained.
5. The urban epidemic spatiotemporal prediction method according to any one of claims 1 to 4, wherein the building of the infectious disease spatiotemporal prediction model based on the graph neural network and the long-short term memory network according to the population movement relationship, the proximity relationship and the position attention relationship comprises:
inputting the graph structure into a graph neural network, wherein the graph neural network normalizes the directed graph by using a neighborhood aggregation method, and the incoming edge weighting of each node is equal to 1:
wherein HiIs a matrix containing the node representations of the previous layer, the initial H0Set as history data representing case changes of infectious diseases in each region, WiAnd f is a nonlinear activation function.
6. The urban epidemic spatiotemporal prediction method according to claim 5, wherein the building of the model for infectious disease spatiotemporal prediction based on the graph neural network and the long-short term memory network according to the population movement relationship, the proximity relationship and the location attention relationship further comprises:
using a message propagation neural network at each time step to obtain a representation sequence hi,t-n,hi,t-n+1,...,hi,t-1Will represent the sequence hi,t-n,hi,t-n+1,...,hi,t-1Inputting the long-term and short-term memory network, and extracting the time sequence relation in the network; the long-short term memory network calculation formula is as follows:
Xi,t=LSTM(hi,t-n,hi,t-n+1,...,hi,t-1)
wherein, Xi,tIs shown asPredicted infectious disease case data of i regions in t time period, hi,t-1Showing the infectious disease case data of the ith area in the t-1 th time slot.
7. The urban epidemic situation spatiotemporal prediction method according to claim 6, wherein the entering of urban infectious disease case data into the infectious disease spatiotemporal prediction model, the obtaining of urban infectious disease prediction results by the infectious disease spatiotemporal prediction model specifically comprises:
fusing a proximity relation, a population movement relation and a position attention relation through an infectious disease space-time prediction model to obtain a city infectious disease prediction result; the fusion mode specifically comprises the following steps:
wherein, Wadj、WodAnd WatIs a matrix of parameters that needs to be trained,andrespectively, the infectious disease prediction result at time t obtained based on the adjacency matrix, the population movement traffic matrix and the location attention matrix, tanh is an activation function,the final prediction result of the whole infectious disease space-time prediction model at the time t is obtained.
8. A city epidemic situation space-time prediction system is characterized by comprising:
a data collection module: the system is used for collecting individual movement track data and infectious disease case data in a city;
a flow calculation module: the system comprises a data driving method, a data acquisition method and a data analysis method, wherein the data driving method is used for processing the individual movement track data, extracting population movement flow among all areas, dividing the population movement flow according to time attributes, and acquiring population movement relations under all the time attributes based on the proximity relations of the areas;
a similarity calculation module: the method comprises the steps of calculating the similarity of infectious disease case data among all regions by using a histogram-based similarity algorithm, and acquiring the position attention relationship of all regions according to the similarity of the infectious disease case data;
a model construction module: the system is used for constructing an infectious disease space-time prediction model based on a graph neural network and a long-short term memory network according to the population movement relationship, the proximity relationship and the position attention relationship;
an infectious disease prediction module: and the system is used for inputting the urban infectious disease case data into the infectious disease space-time prediction model and obtaining the urban infectious disease prediction result through the infectious disease space-time prediction model.
9. A terminal, comprising a processor, a memory coupled to the processor, wherein,
the memory stores program instructions for implementing the urban epidemic spatiotemporal prediction method according to any one of claims 1-7;
the processor is configured to execute the program instructions stored in the memory to control urban epidemic spatiotemporal prediction.
10. A storage medium having stored thereon program instructions executable by a processor to perform the method of urban epidemic spatiotemporal prediction according to any one of claims 1 to 7.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111675147.7A CN114464329A (en) | 2021-12-31 | 2021-12-31 | Urban epidemic situation space-time prediction method, system, terminal and storage medium |
PCT/CN2022/076295 WO2023123625A1 (en) | 2021-12-31 | 2022-02-15 | Urban epidemic space-time prediction method and system, terminal and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111675147.7A CN114464329A (en) | 2021-12-31 | 2021-12-31 | Urban epidemic situation space-time prediction method, system, terminal and storage medium |
Publications (1)
Publication Number | Publication Date |
---|---|
CN114464329A true CN114464329A (en) | 2022-05-10 |
Family
ID=81408260
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111675147.7A Pending CN114464329A (en) | 2021-12-31 | 2021-12-31 | Urban epidemic situation space-time prediction method, system, terminal and storage medium |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN114464329A (en) |
WO (1) | WO2023123625A1 (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116822635A (en) * | 2023-05-12 | 2023-09-29 | 中国科学院深圳先进技术研究院 | Track generation method, device, equipment and storage medium |
CN117150911A (en) * | 2023-09-04 | 2023-12-01 | 吉林建筑大学 | Coal rock instability fracture prediction method and system based on graph neural network |
WO2024031520A1 (en) * | 2022-08-09 | 2024-02-15 | 浙江大学 | Human mobility prediction method based on generative adversarial network |
CN117690601A (en) * | 2024-02-02 | 2024-03-12 | 江西省胸科医院(江西省第三人民医院) | Tuberculosis epidemic trend prediction system based on big data analysis |
CN117711636A (en) * | 2023-12-15 | 2024-03-15 | 南京理工大学 | Monkey pox epidemic prediction method based on tensor space-time diagram convolution of attention mechanism |
CN118471545A (en) * | 2024-07-11 | 2024-08-09 | 中国科学院自动化研究所 | Infectious disease transmission prediction method and device based on data decoupling and electronic equipment |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117332120B (en) * | 2023-08-29 | 2024-04-30 | 泰瑞数创科技(北京)股份有限公司 | Geographic entity relation construction and expression method based on space calculation |
CN117332033B (en) * | 2023-09-27 | 2024-08-16 | 中国科学院深圳先进技术研究院 | Space-time track generation method and device, computer equipment and storage medium |
CN117649028B (en) * | 2024-01-26 | 2024-04-02 | 南京航空航天大学 | Urban function area matching-based inter-urban crowd flow trend prediction method |
CN118314464A (en) * | 2024-06-11 | 2024-07-09 | 华侨大学 | Historical urban area boundary judgment method, system and equipment based on graph neural network |
CN118379175B (en) * | 2024-06-26 | 2024-09-13 | 武汉市测绘研究院 | Urban health toughness space-time characteristic modeling method considering dynamic interaction |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111554408A (en) * | 2020-04-27 | 2020-08-18 | 中国科学院深圳先进技术研究院 | Urban interior dengue space-time prediction method and system and electronic equipment |
CN113689957A (en) * | 2021-08-23 | 2021-11-23 | 北京航空航天大学 | Construction method of fine-grained infectious disease simulation model |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111639787B (en) * | 2020-04-28 | 2024-03-15 | 北京工商大学 | Spatio-temporal data prediction method based on graph convolution network |
CN112102959B (en) * | 2020-11-02 | 2021-03-23 | 平安科技(深圳)有限公司 | Server, data processing method, data processing device and readable storage medium |
CN113314231B (en) * | 2021-05-28 | 2022-04-22 | 北京航空航天大学 | Infectious disease propagation prediction system and device integrating spatio-temporal information |
-
2021
- 2021-12-31 CN CN202111675147.7A patent/CN114464329A/en active Pending
-
2022
- 2022-02-15 WO PCT/CN2022/076295 patent/WO2023123625A1/en unknown
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111554408A (en) * | 2020-04-27 | 2020-08-18 | 中国科学院深圳先进技术研究院 | Urban interior dengue space-time prediction method and system and electronic equipment |
CN113689957A (en) * | 2021-08-23 | 2021-11-23 | 北京航空航天大学 | Construction method of fine-grained infectious disease simulation model |
Non-Patent Citations (2)
Title |
---|
刘明亮: "基于时空网络的城市疫情传播评估与预测", 万方数据知识服务平台, 22 December 2021 (2021-12-22), pages 4 * |
尹凌: "耦合人群移动的COVID-19传染病模型研究进展", 地球信息科学, vol. 23, no. 11, 30 November 2021 (2021-11-30), pages 3 * |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2024031520A1 (en) * | 2022-08-09 | 2024-02-15 | 浙江大学 | Human mobility prediction method based on generative adversarial network |
CN116822635A (en) * | 2023-05-12 | 2023-09-29 | 中国科学院深圳先进技术研究院 | Track generation method, device, equipment and storage medium |
CN117150911A (en) * | 2023-09-04 | 2023-12-01 | 吉林建筑大学 | Coal rock instability fracture prediction method and system based on graph neural network |
CN117150911B (en) * | 2023-09-04 | 2024-04-26 | 吉林建筑大学 | Coal rock instability fracture prediction method and system based on graph neural network |
CN117711636A (en) * | 2023-12-15 | 2024-03-15 | 南京理工大学 | Monkey pox epidemic prediction method based on tensor space-time diagram convolution of attention mechanism |
CN117711636B (en) * | 2023-12-15 | 2024-08-13 | 南京理工大学 | Monkey pox epidemic prediction method based on tensor space-time diagram convolution of attention mechanism |
CN117690601A (en) * | 2024-02-02 | 2024-03-12 | 江西省胸科医院(江西省第三人民医院) | Tuberculosis epidemic trend prediction system based on big data analysis |
CN117690601B (en) * | 2024-02-02 | 2024-05-24 | 江西省胸科医院(江西省第三人民医院) | Tuberculosis epidemic trend prediction system based on big data analysis |
CN118471545A (en) * | 2024-07-11 | 2024-08-09 | 中国科学院自动化研究所 | Infectious disease transmission prediction method and device based on data decoupling and electronic equipment |
Also Published As
Publication number | Publication date |
---|---|
WO2023123625A1 (en) | 2023-07-06 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN114464329A (en) | Urban epidemic situation space-time prediction method, system, terminal and storage medium | |
CN111612206B (en) | Neighborhood people stream prediction method and system based on space-time diagram convolution neural network | |
He et al. | Graph attention spatial-temporal network with collaborative global-local learning for citywide mobile traffic prediction | |
Xiao et al. | Predicting urban region heat via learning arrive-stay-leave behaviors of private cars | |
CN112532439B (en) | Network flow prediction method based on attention multi-component space-time cross-domain neural network model | |
WO2023123624A1 (en) | Method and system for predicting influenza outbreak trend in city, and terminal and storage medium | |
CN113673769B (en) | Traffic flow prediction method of graph neural network based on multivariate time sequence interpolation | |
CN109858390A (en) | The Activity recognition method of human skeleton based on end-to-end space-time diagram learning neural network | |
CN110827544A (en) | Short-term traffic flow control method based on graph convolution recurrent neural network | |
Lv et al. | Blind travel prediction based on obstacle avoidance in indoor scene | |
Alizadeh et al. | Network traffic forecasting based on fixed telecommunication data using deep learning | |
CN113792930B (en) | Blind person walking track prediction method, electronic equipment and storage medium | |
CN106815563B (en) | Human body apparent structure-based crowd quantity prediction method | |
WO2024031520A1 (en) | Human mobility prediction method based on generative adversarial network | |
Kulkarni et al. | Mobidict: A mobility prediction system leveraging realtime location data streams | |
Wan et al. | Addressing location uncertainties in GPS‐based activity monitoring: A methodological framework | |
CN114461931A (en) | User trajectory prediction method and system based on multi-relation fusion analysis | |
CN116502763A (en) | Crowd congestion prediction method and system based on space-time diagram neural network | |
Zhang et al. | A spatiotemporal graph wavelet neural network for traffic flow prediction | |
Wang et al. | A hypergraph-based hybrid graph convolutional network for intracity human activity intensity prediction and geographic relationship interpretation | |
Li et al. | A multi-scale video surveillance based information aggregation model for crime prediction | |
CN116578858A (en) | Air compressor fault prediction and health degree evaluation method and system based on graphic neural network | |
Al-Molegi et al. | Regions-of-interest discovering and predicting in smartphone environments | |
ABBAS | A survey of research into artificial neural networks for crime prediction | |
Jin et al. | Spatial–temporal uncertainty-aware graph networks for promoting accuracy and reliability of traffic forecasting |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |