Font Size: a A A

A Research On Warning Release Model Of Hand,Foot And Mouth Disease In Jiangsu Province Based On Internet Data

Posted on:2020-05-15Degree:MasterType:Thesis
Country:ChinaCandidate:J B TangFull Text:PDF
GTID:2404330620456203Subject:Electronics and Communications Engineering
Abstract/Summary:PDF Full Text Request
Prediction of epidemic diseases using Internet data attracts tremendous attention in epidemiology in recent years.Compared with traditional methods of disease surveillance,disease prediction based on Internet data has the characteristics of low cost and high timeliness.The proposed model in our paper can help decision makers to prevent diseases from further spreading in Jiangsu Province.Our study focuses on the method of applying search engine,social network,weather and air quality data to model the epidemic trend of HFMD(Hand Foot and Mouth Disease)in Jiangsu Province based on machine learning methods,so as to release warnings before the outbreaks.The experimental data in our study consists of Baidu Index,Weibo text,weather data,air quality data and surveillance incidence data,covering 13 prefecture-level cities in Jiangsu Province,spanning the period from 2013 to 2018.The surveillance incidence data is provided from NIDRIS(Notifiable Infectious Diseases Reporting Information System).Our research established a time series prediction model for cities in Jiangsu Province.The first step was to preprocess the Internet data and screen features for model training.The second step was to establish a time series prediction model for the surveillance incidence data based on the screened features.Our proposed time series model was constructed by combining different base models.Then the warning threshold values were calculated according to MEM(Moving Epidemic Method).This study combined the warning threshold values with the time series prediction model to carry out warnings of the outbreaks of HFMD in Jiangsu.Finally,our study established a joint warning model based on NTDDM(Network Threshold Disease Diffusion Model).Our study compares the performance of time series prediction model and joint warning model,and combines the two models to construct the final warning model.The final warning model proposed in this paper is applicable for all the cities in Jiangsu Province in 2018,and able to release warnings 8.8 days in advance.
Keywords/Search Tags:epidemic diseases, internet data, machine learning, Moving Epidemic Method, Network Threshold Disease Diffusion Model
PDF Full Text Request
Related items