Skip to content

Latest commit

 

History

History
17 lines (12 loc) · 1.34 KB

README.md

File metadata and controls

17 lines (12 loc) · 1.34 KB

Time_Series_Analysis - Forecasting of Refugee Inflow via Balkon route

The data set has been attached to the repository if in case you need an online copy follow the below mentioned link : https://data.humdata.org/dataset/daily-estimated-arrivals-through-western-balkans-route 20160921-AnalysisofEstimatedArrivals.xlsx

The notebook has 2 ipynb files

  1. Pre_processing.ipynb - This file contains the coding and results for stationarity check, time series decomposition, making the time series stationary and exporting the count of refugees with eliminated trend and seasonality.

  2. Forecating_model.ipynb - This file contains the coding and results of various forecasting models like AR, ARMA, ARIMA, SARIMA along with it the evaluation results is also available for each model. Finally a short optimisation phase is added which determines the best parameters for the ARIMA model. One can use the same for determining the best parameters for other models by simply changing the calling function.

Note : Since the github has an issue of rendering ipynb files I would recommend you to clone the repo and then load it in your notebook. Else follow the link : https://nbviewer.jupyter.org Then copy the url of the github link of my repo and paste it in the space provided. After which you can view all the cells clearly.

For queries contact: [email protected]