CN112102957A - Epidemic disease early warning method and early warning system thereof - Google Patents

Epidemic disease early warning method and early warning system thereof Download PDF

Info

Publication number
CN112102957A
CN112102957A CN202010911218.8A CN202010911218A CN112102957A CN 112102957 A CN112102957 A CN 112102957A CN 202010911218 A CN202010911218 A CN 202010911218A CN 112102957 A CN112102957 A CN 112102957A
Authority
CN
China
Prior art keywords
epidemic
early warning
purchase rate
historical
sales data
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
Application number
CN202010911218.8A
Other languages
Chinese (zh)
Inventor
罗安
周聪俊
史鹏翔
许春霞
师改梅
何进
徐�明
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Sichuan Junyi Fudun Technology Co ltd
Original Assignee
Sichuan Junyi Fudun Technology Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Sichuan Junyi Fudun Technology Co ltd filed Critical Sichuan Junyi Fudun Technology Co ltd
Priority to CN202010911218.8A priority Critical patent/CN112102957A/en
Publication of CN112102957A publication Critical patent/CN112102957A/en
Pending legal-status Critical Current

Links

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
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • 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"
    • 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
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0201—Market modelling; Market analysis; Collecting market data
    • G06Q30/0202—Market predictions or forecasting for commercial activities
    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
    • G08B21/02—Alarms for ensuring the safety of persons

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Strategic Management (AREA)
  • General Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Development Economics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Finance (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Software Systems (AREA)
  • Molecular Biology (AREA)
  • Public Health (AREA)
  • Computing Systems (AREA)
  • Evolutionary Computation (AREA)
  • Accounting & Taxation (AREA)
  • Economics (AREA)
  • Mathematical Physics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • General Engineering & Computer Science (AREA)
  • General Business, Economics & Management (AREA)
  • Marketing (AREA)
  • Game Theory and Decision Science (AREA)
  • Human Resources & Organizations (AREA)
  • Medical Informatics (AREA)
  • Databases & Information Systems (AREA)
  • Epidemiology (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • Primary Health Care (AREA)
  • Pathology (AREA)
  • Emergency Management (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention belongs to the technical field of epidemic disease early warning, and particularly relates to an epidemic disease early warning method and an early warning system thereof, wherein the epidemic disease early warning method comprises the following steps: s1: acquiring historical drug sales data, and preprocessing the historical drug sales data; s2: obtaining the historical purchase rate of epidemic medicines according to the preprocessed historical medicine sales data; s3: establishing a medicine sales prediction model based on a neural network according to the historical purchase rate of epidemic medicines; s4: acquiring real-time medicine sales data, and predicting by using a medicine sales prediction model according to the real-time medicine sales data to obtain the predicted purchase rate of epidemic medicines; s5: and judging whether the predicted purchase rate of the epidemic medicines is abnormal or not, and if so, sending out an epidemic early warning signal. The invention solves the problems that the prior art can not meet the requirements of epidemic disease monitoring and lacks an analysis and prediction method for medicine sales data.

Description

Epidemic disease early warning method and early warning system thereof
Technical Field
The invention belongs to the technical field of epidemic disease early warning, and particularly relates to an epidemic disease early warning method and an early warning system thereof.
Background
In recent years, with the emergence of new infectious diseases and the scarcity of existing infectious diseases (influenza, pulmonary tuberculosis, meningitis, cryptosporidiosis and the like), the traditional disease monitoring system based on diagnosis cannot meet the requirement of emergency treatment of emergent public health events, and the symptom monitoring (syndrome surveillance) can make up the defect to a certain extent. Symptom monitoring is a mechanism for detecting, evaluating and reporting possible emergencies, mainly early detection and investigation of disease occurrence, and is a research hotspot developed in recent years for coping with biological and chemical terrorist attacks or other serious public health event hazards.
The occurrence and development of diseases are complex phenomena, and the accurate prediction of the development trend of people and individual diseases becomes an important means for preventing diseases. As one of the symptom monitors, the medicine sales monitor has been started to be applied abroad. By continuously and systematically collecting, checking and analyzing drug sales and influencing factor data, the disease onset trend of related diseases is predicted, and epidemic outbreak and public health events are early warned.
Drug sales monitoring is a symptom monitoring based on the sales of retail drugs in drugstores, and the theoretical hypothesis is that patients tend to purchase drugs in drugstores for self-treatment in the early stage of diseases, thereby causing the sales of drugs to increase, so that the monitoring of drug sales may discover the existence of diseases earlier than the conventional monitoring system. The sales volume of the daily retail medicines of the pharmacy is tracked and monitored to estimate and track the scale and the trend of disease outbreak, and possible public health incidents are early warned to remind people to take effective measures in time, so that the morbidity and the mortality are reduced, and the economic loss is reduced.
Therefore, by monitoring the change in the sales amount of the medicine and determining the threshold value, it is theoretically possible to find information on the presence of the accumulation of the disease earlier than the conventional monitoring method. However, not all infectious diseases are suitable for early warning by drug sales monitoring, and are generally suitable for influenza and other epidemic diseases which mainly cause dominant infection, have more mild patients, have larger epidemic scale, have no or few specific symptoms and are known by residents to a little of treatment attempts.
The sales prediction is a prediction of future sales based on a conventional sales situation using a sales prediction model. At present, most sales volume prediction is carried out by adopting a method based on time series analysis, the time series analysis method uses past historical data and further predicts the future development trend through statistical analysis, and the basis is that all things are developed and changed and the development and change of the things are continuous in time. Due to the complexities associated with the medical field, methods of analysis and prediction for drug sales data are lacking.
In the process of implementing the invention, the inventor finds that at least the following problems exist in the prior art:
in the prior art, a traditional disease monitoring system based on diagnosis cannot meet the requirement of emergency treatment of emergent public health events and can not meet the requirement of epidemic disease monitoring, most of the current sales volume prediction adopts a method based on time series analysis to predict, and an analysis and prediction method aiming at medicine sales data is lacked.
Disclosure of Invention
The present invention aims to solve at least one of the above technical problems to a certain extent.
Therefore, the invention aims to provide an epidemic early warning method and an epidemic early warning system, which are used for solving the problems that the prior art cannot meet the requirement of epidemic disease monitoring and lacks of an analysis and prediction method for medicine sales data.
The technical scheme adopted by the invention is as follows:
an epidemic early warning method comprises the following steps:
s1: acquiring historical drug sales data, and preprocessing the historical drug sales data;
s2: obtaining the historical purchase rate of epidemic medicines according to the preprocessed historical medicine sales data;
s3: establishing a medicine sales prediction model based on a neural network according to the historical purchase rate of epidemic medicines;
s4: acquiring real-time medicine sales data, and predicting by using a medicine sales prediction model according to the real-time medicine sales data to obtain the predicted purchase rate of epidemic medicines;
s5: and judging whether the predicted purchase rate of the epidemic medicines is abnormal or not, if so, sending out an epidemic early warning signal, and ending the epidemic early warning method, otherwise, returning to the step S4.
Further, the historical drug sales data comprises sales data of all drug types within a preset historical time limit;
the real-time drug sales data includes sales data for all drug categories within a preset period.
Further, the specific method for acquiring the purchase rate of epidemic medicines comprises the following steps: and dividing the sales data of the current epidemic medicines by the sales data of all the medicine types in the current day to obtain the purchase rate of the current epidemic medicines.
Further, in step S1, the preprocessing includes an outlier analysis processing and a normalization processing.
Further, in step S3, the medicine sales prediction model is a BP neural network combined prediction model, and the BP neural network combined prediction model is formed by combining a first combined model and a second combined model.
Further, the first combination model is formed by combining a gray prediction model and an LSTM prediction model;
the second combined model is formed by combining an ARIMA model and an LSTM prediction model.
Further, the formula of the first combination model is:
Figure DEST_PATH_IMAGE001
in the formula (I), the compound is shown in the specification,
Figure 394421DEST_PATH_IMAGE002
for the first combined model atN+mA predicted purchase rate of a drug for an epidemic;
Figure DEST_PATH_IMAGE003
for the gray prediction model in the first combined modelN+mA daily forecast purchase rate;
Figure 639457DEST_PATH_IMAGE004
for the LSTM prediction model in the first combined model in the secondN+mPurchase rate residuals for days;Ntotal days in historical years;mamounts are indicated for the predicted days.
Further, the formula of the second combination model is:
Figure DEST_PATH_IMAGE005
in the formula (I), the compound is shown in the specification,
Figure 950353DEST_PATH_IMAGE006
for the second combined model inN+mA predicted purchase rate of a drug for an epidemic;
Figure DEST_PATH_IMAGE007
for ARIMA model in the second combined model at the firstN+mA daily forecast purchase rate;
Figure 611141DEST_PATH_IMAGE008
for the LSTM prediction model in the second combined model in the firstN+mA daily purchase rate residual;Ntotal days in historical years;mamounts are indicated for the predicted days.
Further, in step S5, the abnormality determination formula for the predicted purchase rate of epidemic drugs is:
Figure DEST_PATH_IMAGE009
in the formula (I), the compound is shown in the specification,
Figure 753410DEST_PATH_IMAGE010
the actual purchase rate of the current day epidemic medicines is obtained according to the real-time medicine sales data;
Figure DEST_PATH_IMAGE011
the predicted purchase rate of the current day epidemic medicine is obtained according to the medicine sales prediction model;this the purchase rate threshold.
An epidemic disease early warning system is applied to an epidemic disease early warning method, and comprises a data acquisition and preprocessing unit, a sales prediction unit and an epidemic disease early warning unit which are sequentially in communication connection;
the data acquisition and preprocessing unit is used for acquiring historical medicine sales data and preprocessing the historical medicine sales data;
the sales prediction unit is used for obtaining the historical purchase rate of epidemic medicines according to the preprocessed historical medicine sales data, establishing a medicine sales prediction model based on a neural network according to the historical purchase rate of the epidemic medicines, obtaining real-time medicine sales data, predicting by using the medicine sales prediction model according to the real-time medicine sales data to obtain the predicted purchase rate of the epidemic medicines, judging whether the predicted purchase rate of the epidemic medicines is abnormal or not, if so, controlling the epidemic early warning unit to send out an epidemic early warning signal, ending the epidemic early warning method, and if not, obtaining the real-time medicine sales data, and performing subsequent sales prediction tasks;
and the epidemic disease early warning unit is used for sending out an epidemic disease early warning signal.
The invention has the beneficial effects that:
the epidemic early warning method is based on analysis of sales data of epidemic medicines, and a medicine sales prediction model is established to predict future epidemic medicine sales data, so that the epidemic disease monitoring efficiency and accuracy are improved, early warning signals are rapidly found and sent out in the early stage of epidemic diseases, and occurrence and development of emergent public health events are effectively prevented and controlled.
Other advantageous effects of the present invention will be described in detail in the detailed description.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a flow chart of a method of epidemic early warning;
fig. 2 is a block diagram of the epidemic early warning system.
Detailed Description
The invention is further described with reference to the following figures and specific embodiments. It should be noted that the description of the embodiments is provided to help understanding of the present invention, but the present invention is not limited thereto. Functional details disclosed herein are merely illustrative of example embodiments of the invention. This invention may, however, be embodied in many alternate forms and should not be construed as limited to the embodiments set forth herein.
It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments of the invention. When the terms "comprises," "comprising," "includes," and/or "including" are used herein, they specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, numbers, steps, operations, elements, components, and/or groups thereof.
It should also be noted that, in some alternative implementations, the functions/acts noted may occur out of the order noted in the figures. For example, two figures shown in succession may, in fact, be executed substantially concurrently, or the figures may sometimes be executed in the reverse order, depending upon the functionality/acts involved.
It should be understood that specific details are provided in the following description to facilitate a thorough understanding of example embodiments. However, it will be understood by those of ordinary skill in the art that the example embodiments may be practiced without these specific details. For example, systems may be shown in block diagrams in order not to obscure the examples in unnecessary detail. In other instances, well-known processes, structures and techniques may be shown without unnecessary detail in order to avoid obscuring example embodiments.
Example 1
As shown in fig. 1, the embodiment provides an epidemic early warning method, which includes the following steps:
s1: acquiring historical drug sales data, and preprocessing the historical drug sales data;
preprocessing comprises abnormal value analysis processing and normalization processing;
s2: obtaining the historical purchase rate of epidemic medicines according to the preprocessed historical medicine sales data;
s3: establishing a medicine sales prediction model based on a neural network according to the historical purchase rate of epidemic medicines;
the drug sales prediction model is a BP neural network combined prediction model, the BP neural network combined prediction model is formed by combining a first combined model and a second combined model, and the problem that the prediction precision is not high due to a single prediction model when the drug sales prediction model is over-fitted is prevented;
the first combination model is formed by combining a gray prediction model and a long-time memory network LSTM prediction model;
the method comprises the following specific steps:
a-1: predicting the purchase rate of the epidemic disease medicines by using a gray prediction model of the medicine sales prediction model;
1) based on historical purchase rate data for epidemic medications
Figure 841452DEST_PATH_IMAGE012
Accumulating to obtain a historical purchase rate fitting sequence of epidemic medicines
Figure DEST_PATH_IMAGE013
Wherein, in the step (A),
Figure 639643DEST_PATH_IMAGE014
for the historical purchase rate of the current day's epidemic drugs,
Figure DEST_PATH_IMAGE015
accumulated value of historical purchase rate of epidemic drugs on the same day,Ntotal days in historical years;
2) according to historical purchase rate data of epidemic medicines, establishing an intermediate matrix:
the formula of the intermediate matrix is:
Figure 166440DEST_PATH_IMAGE016
Figure DEST_PATH_IMAGE017
in the formula (I), the compound is shown in the specification,Bis a first intermediate matrix;
Figure 100897DEST_PATH_IMAGE018
is a second intermediate matrix;
Figure DEST_PATH_IMAGE019
is the historical rate of purchase of the current day's epidemic drugs;
Figure 625420DEST_PATH_IMAGE020
historical purchase rate of epidemic drugs for the day for which the grey prediction model was entered;
3) obtaining an estimated value according to the intermediate matrix;
Figure DEST_PATH_IMAGE021
in the formula (I), the compound is shown in the specification,
Figure 707645DEST_PATH_IMAGE022
obtaining a matrix for the estimated value;Bis a first intermediate matrix;
Figure DEST_PATH_IMAGE023
is a second intermediate matrix;
Figure 716096DEST_PATH_IMAGE024
is a first intermediate matrixBThe transposed matrix of (2);
Figure DEST_PATH_IMAGE025
is a first estimated value;
Figure 770639DEST_PATH_IMAGE026
is a second estimated value;
4) establishing a gray prediction model according to the estimation value;
the formula for the gray prediction model is:
Figure DEST_PATH_IMAGE027
in the formula (I), the compound is shown in the specification,
Figure 528380DEST_PATH_IMAGE028
as a grey prediction model inn+1 day predicted purchase rate of epidemic drugs;
Figure DEST_PATH_IMAGE029
historical purchase rate of epidemic drugs on day 1 for the grey prediction model;
Figure 301164DEST_PATH_IMAGE025
is a first estimated value;
Figure 107446DEST_PATH_IMAGE030
is a second estimated value;nindicated amounts for days;
5) acquiring the predicted purchase rate of epidemic medicines according to the grey prediction model;
obtaining fitting values using a gray prediction model
Figure DEST_PATH_IMAGE031
Fitting values according to a post-subtraction operation
Figure 16496DEST_PATH_IMAGE032
Is subjected to a reduction treatment, i.e.
Figure DEST_PATH_IMAGE033
，
Figure 882821DEST_PATH_IMAGE034
As a grey prediction model ini、i+1 day predicted purchase rate of epidemic drugs;
obtained according to a grey prediction model
Figure DEST_PATH_IMAGE035
Based on the post subtraction operation, obtainnPredicted value of purchase rate of epidemic medicine in +1 day
Figure 939638DEST_PATH_IMAGE036
，
Figure DEST_PATH_IMAGE037
Obtaining a purchase rate fitting sequence of historical epidemic medicines
Figure 549611DEST_PATH_IMAGE038
And predicted purchase rate of epidemic drugs
Figure DEST_PATH_IMAGE039
；
6) Fitting sequence of historical purchase rate of epidemic drugs obtained by grey prediction model
Figure 313168DEST_PATH_IMAGE040
Historical purchase rate data associated with epidemic medications
Figure DEST_PATH_IMAGE041
Subtracting to obtain the historical purchase rate residual error sequence of epidemic drugs
Figure 615973DEST_PATH_IMAGE042
；
A-2: predicting the purchase rate residual error sequence of the epidemic disease medicines by using an LSTM prediction model according to the historical purchase rate residual error sequence of the epidemic disease medicines;
1) historical purchase rate residual error sequence of epidemic medicine obtained by grey prediction model
Figure DEST_PATH_IMAGE043
Carrying out normalization processing;
2) constructing a multilayer LSTM model, and training to obtain an optimal LSTM prediction model;
adopting a 4-layer LSTM neural network model, adopting a RELU (equal weighted average) activation function as an activation function, adopting a mean square error function as a loss function, adopting an adam Aadm optimization algorithm as an optimizer, and adopting a random inactivation Dropout method to prevent overfitting, wherein the first hidden layer and the second hidden layer both adopt 30 neurons, and the adopted Dropout proportion is 0.2;
3) adopting a rolling prediction method, and obtaining a predicted purchase rate residual error sequence of epidemic medicines by using an optimal LSTM prediction model
Figure 894508DEST_PATH_IMAGE044
；
A-3: combining the grey prediction model with the LSTM prediction model to obtain a first combination model;
the formula of the first combined model is:
Figure DEST_PATH_IMAGE045
in the formula (I), the compound is shown in the specification,
Figure 42593DEST_PATH_IMAGE046
for the first combined model atN+mA predicted purchase rate of a drug for an epidemic;
Figure 191814DEST_PATH_IMAGE003
for the gray prediction model in the first combined modelN+mA daily forecast purchase rate;
Figure 868783DEST_PATH_IMAGE004
for the LSTM prediction model in the first combined model in the secondN+mPurchase rate residuals for days;Ntotal days in historical years;min order to predict the indicated amount of days,
Figure DEST_PATH_IMAGE047
from this, the predicted purchase rates of epidemic drugs can be derived for the first day of the month, the second day of the month, and the thirty-th day:
Figure 634614DEST_PATH_IMAGE048
；
the second combined model is formed by combining a difference integration moving average autoregressive ARIMA model and an LSTM prediction model;
the method comprises the following specific steps:
b-1: predicting the purchase rate of the epidemic disease medicine by using an ARIMA model of a medicine sales prediction model;
1) determining historical purchase rate data for epidemic drugs
Figure DEST_PATH_IMAGE049
If so, entering the step 3), otherwise, entering the step 2);
2) historical purchase rate number of epidemic disease medicineAccording to the proceedingdThe secondary difference smoothes the historical rate of purchase of epidemic drugs, which, in embodiments of the invention,
Figure 320810DEST_PATH_IMAGE050
；
3) carrying out parameter solution on the ARIMA model, and establishing the ARIMA model;
determining the order approximately through an autocorrelation EACF function, considering the understanding of the size of the Chichi cell information content AIC, finally determining the order of the model, performing parameter estimation after determining the order of the model, and checking the significance of the parameters and the rationality of the model;
4) obtaining a historical purchase rate fitting sequence of epidemic drugs according to an ARIMA model
Figure DEST_PATH_IMAGE051
And predicted purchase rate residual sequence of epidemic drugs
Figure 324538DEST_PATH_IMAGE052
；
5) Fitting historical purchase rates of epidemic drugs to a sequence
Figure DEST_PATH_IMAGE053
Historical purchase rate sequence with epidemic drugs
Figure 765884DEST_PATH_IMAGE054
Subtracting to obtain the historical purchase rate residual error sequence of epidemic drugs
Figure DEST_PATH_IMAGE055
；
6) Judging whether the maximum iteration times is reached or the error requirement is met, if so, returning to the step 2), and otherwise, entering the step 7);
7) adjusting parameters of the ARIMA model, returning to the step 3), and performing iteration;
b-2: predicting the purchase rate residual error sequence of the epidemic disease medicines by using an LSTM prediction model according to the historical purchase rate residual error sequence of the epidemic disease medicines;
adopting a rolling prediction method, and obtaining a predicted purchase rate residual error sequence of epidemic medicines by using an optimal LSTM prediction model
Figure 956694DEST_PATH_IMAGE056
；
B-3: combining the ARIMA model with the LSTM prediction model to obtain a first combination model;
the formula of the second combined model is:
Figure 977739DEST_PATH_IMAGE005
in the formula (I), the compound is shown in the specification,
Figure DEST_PATH_IMAGE057
for the second combined model inN+mA predicted purchase rate of a drug for an epidemic;
Figure 304815DEST_PATH_IMAGE058
for ARIMA model in the second combined model at the firstN+mA daily forecast purchase rate;
Figure 385904DEST_PATH_IMAGE059
for the LSTM prediction model in the second combined model in the firstN+mA daily purchase rate residual;Ntotal days in historical years;min order to predict the indicated amount of days,
Figure DEST_PATH_IMAGE060
from this, the predicted purchase rates of epidemic drugs can be derived for the first day of the month, the second day of the month, and the thirty-th day:
Figure 64010DEST_PATH_IMAGE061
；
the single model prediction result is used as the input of a BP neural network, the prediction purchase rate of epidemic medicines is used as the output, and the combined weight can be obtained through 3-layer BP network learning and training according to the criterion of minimum mean square error;
s4: acquiring real-time medicine sales data, and predicting by using a medicine sales prediction model according to the real-time medicine sales data to obtain the predicted purchase rate of epidemic medicines;
according to the invention, a rolling sales prediction method is adopted, real-time medicine sales are brought into prediction consideration regularly to respond to actual product demands of the market, and the future prediction data are adjusted according to actual sales conditions of the market as soon as new epidemic medicine purchase rate data are generated monthly, so that the prediction accuracy is improved, epidemic situation changes can be perceived more quickly, and more comprehensive decision consideration is made;
s5: judging whether the predicted purchase rate of the epidemic medicines is abnormal or not, if so, sending out an epidemic early warning signal, and ending the epidemic early warning method, otherwise, returning to the step S4;
the abnormal judgment formula of the predicted purchase rate of the epidemic medicines is as follows:
Figure 888747DEST_PATH_IMAGE062
in the formula (I), the compound is shown in the specification,
Figure DEST_PATH_IMAGE063
the actual purchase rate of the current day epidemic medicines is obtained according to the real-time medicine sales data;
Figure 335908DEST_PATH_IMAGE011
the predicted purchase rate of the current day epidemic medicine is obtained according to the medicine sales prediction model;this the purchase rate threshold.
Preferably, the historical drug sales data comprises sales data of all drug types within a preset historical time limit;
the real-time drug sales data includes sales data for all drug categories within a preset period.
Preferably, the specific method for acquiring the purchase rate of the epidemic medicine comprises the following steps: and dividing the sales data of the current epidemic medicines by the sales data of all the medicine types in the current day to obtain the purchase rate of the current epidemic medicines.
The epidemic early warning method is based on analysis of sales data of epidemic medicines, and a medicine sales prediction model is established to predict future epidemic medicine sales data, so that the epidemic disease monitoring efficiency and accuracy are improved, early warning signals are rapidly found and sent out in the early stage of epidemic diseases, and occurrence and development of emergent public health events are effectively prevented and controlled.
As shown in fig. 2, an epidemic early warning system is applied to an epidemic early warning method, and includes a data acquisition and preprocessing unit, a sales prediction unit, and an epidemic early warning unit, which are sequentially connected in communication;
the data acquisition and preprocessing unit is used for acquiring historical medicine sales data and preprocessing the historical medicine sales data;
the sales prediction unit is used for obtaining the historical purchase rate of epidemic medicines according to the preprocessed historical medicine sales data, establishing a medicine sales prediction model based on a neural network according to the historical purchase rate of the epidemic medicines, obtaining real-time medicine sales data, predicting by using the medicine sales prediction model according to the real-time medicine sales data to obtain the predicted purchase rate of the epidemic medicines, judging whether the predicted purchase rate of the epidemic medicines is abnormal or not, if so, controlling the epidemic early warning unit to send out an epidemic early warning signal, ending the epidemic early warning method, and if not, obtaining the real-time medicine sales data, and performing subsequent sales prediction tasks;
and the epidemic disease early warning unit is used for sending out an epidemic disease early warning signal.
It will be apparent to those skilled in the art that the modules or steps of the present invention described above may be implemented by a general purpose computing device, they may be centralized on a single computing device or distributed across a network of multiple computing devices, and they may alternatively be implemented by program code executable by a computing device, such that they may be stored in a storage device and executed by a computing device, or fabricated separately as individual integrated circuit modules, or fabricated as a single integrated circuit module from multiple modules or steps. Thus, the present invention is not limited to any specific combination of hardware and software.
The embodiments described above are merely illustrative, and may or may not be physically separate, if referring to units illustrated as separate components; if reference is made to a component displayed as a unit, it may or may not be a physical unit, and may be located in one place or distributed over a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
The above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: modifications of the technical solutions described in the embodiments or equivalent replacements of some technical features may still be made. And such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
The present invention is not limited to the above-described alternative embodiments, and various other forms of products can be obtained by anyone in light of the present invention. The above detailed description should not be taken as limiting the scope of the invention, which is defined in the claims, and which the description is intended to be interpreted accordingly.

Claims (10)

1. An epidemic disease early warning method is characterized in that: the method comprises the following steps:
s1: acquiring historical drug sales data, and preprocessing the historical drug sales data;
s2: obtaining the historical purchase rate of epidemic medicines according to the preprocessed historical medicine sales data;
s3: establishing a medicine sales prediction model based on a neural network according to the historical purchase rate of epidemic medicines;
s4: acquiring real-time medicine sales data, and predicting by using a medicine sales prediction model according to the real-time medicine sales data to obtain the predicted purchase rate of epidemic medicines;
s5: and judging whether the predicted purchase rate of the epidemic medicines is abnormal or not, if so, sending out an epidemic early warning signal, and ending the epidemic early warning method, otherwise, returning to the step S4.
2. The epidemic early warning method of claim 1, wherein: the historical drug sales data comprises sales data of all drug types within a preset historical time limit;
the real-time drug sales data comprises sales data of all drug types in a preset period.
3. The epidemic early warning method of claim 2, wherein: the specific method for acquiring the purchase rate of epidemic medicines comprises the following steps: and dividing the sales data of the current epidemic medicines by the sales data of all the medicine types in the current day to obtain the purchase rate of the current epidemic medicines.
4. The epidemic early warning method of claim 1, wherein: in step S1, the preprocessing includes an outlier analysis processing and a normalization processing.
5. The epidemic early warning method of claim 1, wherein: in step S3, the drug sales prediction model is a BP neural network combined prediction model, and the BP neural network combined prediction model is formed by combining a first combined model and a second combined model.
6. The epidemic early warning method of claim 5, wherein: the first combination model is formed by combining a gray prediction model and an LSTM prediction model;
the second combined model is formed by combining an ARIMA model and an LSTM prediction model.
7. The epidemic early warning method of claim 6, wherein: the formula of the first combination model is as follows:
Figure DEST_PATH_IMAGE002
in the formula (I), the compound is shown in the specification,
Figure DEST_PATH_IMAGE004
for the first combined model atN+mA predicted purchase rate of a drug for an epidemic;
Figure DEST_PATH_IMAGE006
for the gray prediction model in the first combined modelN+mA daily forecast purchase rate;
Figure DEST_PATH_IMAGE008
for the LSTM prediction model in the first combined model in the secondN+mPurchase rate residuals for days;Ntotal days in historical years;mamounts are indicated for the predicted days.
8. The epidemic early warning method of claim 6, wherein: the formula of the second combination model is as follows:
Figure DEST_PATH_IMAGE010
in the formula (I), the compound is shown in the specification,
Figure DEST_PATH_IMAGE012
for the second combined model inN+mA predicted purchase rate of a drug for an epidemic;
Figure DEST_PATH_IMAGE014
is a second groupThe ARIMA model in the synthetic model isN+mA daily forecast purchase rate;
Figure DEST_PATH_IMAGE016
for the LSTM prediction model in the second combined model in the firstN+mA daily purchase rate residual;Ntotal days in historical years;mamounts are indicated for the predicted days.
9. The epidemic early warning method of claim 1, wherein: in step S5, the abnormality determination formula for the predicted purchase rate of the epidemic medicine is:
Figure DEST_PATH_IMAGE018
in the formula (I), the compound is shown in the specification,
Figure DEST_PATH_IMAGE020
the actual purchase rate of the current day epidemic medicines is obtained according to the real-time medicine sales data;
Figure DEST_PATH_IMAGE022
the predicted purchase rate of the current day epidemic medicine is obtained according to the medicine sales prediction model;this the purchase rate threshold.
10. An epidemic early warning system, which is applied to the epidemic early warning method of any one of claims 1 to 9, and is characterized in that: the epidemic disease early warning system comprises a data acquisition and preprocessing unit, a sales prediction unit and an epidemic disease early warning unit, wherein the data acquisition and preprocessing unit, the sales prediction unit and the epidemic disease early warning unit are sequentially in communication connection;
the data acquisition and preprocessing unit is used for acquiring historical medicine sales data and preprocessing the historical medicine sales data;
the sales prediction unit is used for obtaining the historical purchase rate of epidemic medicines according to the preprocessed historical medicine sales data, establishing a medicine sales prediction model based on a neural network according to the historical purchase rate of the epidemic medicines, obtaining real-time medicine sales data, predicting by using the medicine sales prediction model according to the real-time medicine sales data to obtain the predicted purchase rate of the epidemic medicines, judging whether the predicted purchase rate of the epidemic medicines is abnormal or not, if so, controlling the epidemic early warning unit to send out an epidemic early warning signal, ending the epidemic early warning method, and if not, obtaining the real-time medicine sales data, and performing subsequent sales prediction tasks;
the epidemic disease early warning unit is used for sending out an epidemic disease early warning signal.
CN202010911218.8A 2020-09-02 2020-09-02 Epidemic disease early warning method and early warning system thereof Pending CN112102957A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010911218.8A CN112102957A (en) 2020-09-02 2020-09-02 Epidemic disease early warning method and early warning system thereof

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010911218.8A CN112102957A (en) 2020-09-02 2020-09-02 Epidemic disease early warning method and early warning system thereof

Publications (1)

Publication Number Publication Date
CN112102957A true CN112102957A (en) 2020-12-18

Family

ID=73757154

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010911218.8A Pending CN112102957A (en) 2020-09-02 2020-09-02 Epidemic disease early warning method and early warning system thereof

Country Status (1)

Country Link
CN (1) CN112102957A (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113409952A (en) * 2021-08-20 2021-09-17 苏州市疾病预防控制中心 Infectious disease monitoring, prevention and control system and method under multi-point trigger view angle
CN116110602A (en) * 2023-04-13 2023-05-12 云南医无界医疗网络科技有限公司 Information processing method and system applied to medical community
CN116703528A (en) * 2023-07-31 2023-09-05 山东资略信息技术有限公司 Medical sales management system and management method thereof
CN117095831A (en) * 2023-10-17 2023-11-21 厦门畅享信息技术有限公司 Method, system, medium and electronic equipment for monitoring sudden epidemic trend

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113409952A (en) * 2021-08-20 2021-09-17 苏州市疾病预防控制中心 Infectious disease monitoring, prevention and control system and method under multi-point trigger view angle
CN113409952B (en) * 2021-08-20 2021-11-26 苏州市疾病预防控制中心 Infectious disease monitoring, prevention and control system and method under multi-point trigger view angle
CN116110602A (en) * 2023-04-13 2023-05-12 云南医无界医疗网络科技有限公司 Information processing method and system applied to medical community
CN116110602B (en) * 2023-04-13 2023-06-20 云南医无界医疗网络科技有限公司 Information processing method and system applied to medical community
CN116703528A (en) * 2023-07-31 2023-09-05 山东资略信息技术有限公司 Medical sales management system and management method thereof
CN116703528B (en) * 2023-07-31 2023-11-17 山东资略信息技术有限公司 Medical sales management system and management method thereof
CN117095831A (en) * 2023-10-17 2023-11-21 厦门畅享信息技术有限公司 Method, system, medium and electronic equipment for monitoring sudden epidemic trend
CN117095831B (en) * 2023-10-17 2024-01-16 厦门畅享信息技术有限公司 Method, system, medium and electronic equipment for monitoring sudden epidemic trend

Similar Documents

Publication Publication Date Title
CN112102957A (en) Epidemic disease early warning method and early warning system thereof
CN108197845A (en) A kind of monitoring method of the transaction Indexes Abnormality based on deep learning model LSTM
Lombardo et al. ESSENCE II and the framework for evaluating syndromic surveillance systems.
US20190167103A1 (en) Integrated Sensor Network Methods and Systems
US20210241923A1 (en) Sensor-based machine learning in a health prediction environment
CN109166630B (en) Infectious disease data monitoring and processing method and system
CN103716180A (en) Network flow actual forecasting-based network abnormality pre-warning method
CN1564219A (en) Household security device for lonely living edged people
CN113555136B (en) Five-in-one comprehensive medical and nutritional system based on medical and industrial fusion technology
CN106104539A (en) Use the optimization that the alarm for alarm consulting of alarm regeneration is arranged
US11432778B2 (en) Methods and systems for patient monitoring
US20230118182A1 (en) Remote Monitoring With Artificial Intelligence And Awareness Machines
CN115662631A (en) AI intelligence discrimination-based nursing home management system
Kleinman et al. Simulation for assessing statistical methods of biologic terrorism surveillance.
CN114758786A (en) Dynamic early warning system for post-traumatic hemorrhagic shock based on noninvasive parameters
Brownsell et al. Future systems for remote health care
CN116936134B (en) Complications monitoring method and system based on nursing morning shift data
US20060069535A1 (en) System and method for converting data, and system and method for providing warning signals
Wang Computer internet of things-based intelligent medical system to be applied in home care of senile dementia patients
CN114782234A (en) Intelligent household health management system and method
CN113593680A (en) Hospital congestion mitigation strategy evaluation method
CN117271994B (en) Door detection method and system for care monitoring of old people
Nelson et al. Comparative study of the effects of type I and type II diabetes on human health using statistical models
Plate et al. Predicting clinical deterioration at the intermediate care unit: Comparing a joint modelling approach with a long short-term recurrent neural network
US20110145183A1 (en) Predictor of the period of psychotic episode in individual schizophrenics and its method

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