Skip to content

My senior year project. Forecasting Turkey's real GDP growth using mixed data sampling regression model.

License

Notifications You must be signed in to change notification settings

kazimsanlav/Midas-GDP-Forecast

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

15 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Midas-GDP-Forecast

DOI

My senior project; Forecasting Turkey real GDP growth using mixed frequency data sampling regression model. Based on a recent paper on Journal of Statistical Software, I applied mixed data frequency regression to our problem using R. Also I made an connection between R and Bloomberg api to transfer economitric datas directly to my model in R. In terms of data selection(from bloomberg) and model creation(by created helper function) this model remains flexible and easy to interpret. I believe this work can be used for government and companies who need accurate gdp forecast. More importantly, concept of midas regression can be applied to all sort of problems where different frequency time-series data needed for the model. Link to final presentation

Not: You can use this study in your own work but please refer back to this study.

Final real GDP Growth Prediction

alt text

Nowcast Accuracy

alt text

Nowcast Error Statistic

alt text

About

My senior year project. Forecasting Turkey's real GDP growth using mixed data sampling regression model.

Topics

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages