Stochastic Prediction Model, Case of the Dengue outbreak at Tainan, 2015

  • R1
  • 第 2 天,11:30‑12:15
  • 中文演講/英文投影片
  • 資料分析
  • 進階
  • 不錄影

Statistical models play an important role in disease prediction. Time serialized incidence data can be used to predict the future occurrence of disease events. In the case of modeling the Taiwan's dengue outbreak, 2015, we should discuss how to refine/manipulate the data to be functioned and provide an opportunity to develop some stochastic models for predictive power.


Since 2015, dengue out-break damaged Tainan area and strongly affected people's life very much. This brings the experience that the implict threaten of dieases could be not blocked due to the space limit as usual and it is important to moniter epidemiological inference and make right public health planning just in time. In this talk, we will propose a detailed procedure of building up monitoring and forcast system, including 1. Data clean/manipulation for use, - Infomation visualisation, statistical results and geolocation related - setup the monitoring system by estimation of the attack ratio of an epidemic, $R_t$. - building stochastic models for the prediction of infectious disease outbreaks. **Note**. cross-comparison with other third-party module would be mentioned, """fbprophet""" for instance. Python and required modules include: numpy, scipy, matplotlib, pandas, gmplot, gqrid, JSAnimation, seaborn, statmodels, arch, ipywidgets-6.0.0 which are selected under the concern: easy-to-installation. Case studies will run on Python Jupyter notebook environment (Python 3.6).



Chu-Ching Huang

A python user
Chang-Gung University, Center of Education


Undergraduate in NCTU.
Majoring in financing.