It is a 5-year project to research emergency visits stream distribution geographically and temporally. Data are from emergencies at Île-de-France from 2014 to 2015.
Firstly, data cleaning is done in order to better understand all meanings of variables (62 variables).
Next, statistics analyses present emergency visits stream's characteristics to the extend of genders, ages, time, festivals and so on.
Finally, three models are proposed to forecast emergency visits: a benchmark (mean visits per 6 hours), a linear regression and Sarima model.
- This project is confidential, so no graph or data is presented in this repository.