Skip to content

angyurchenko/Prophet-forecast

Repository files navigation

Prophet-forecast

Summary:

Write a program in Python to run prediction with the Prophet algorithm (developed by FB team)

There are two files with test data attached: past.csv and future.csv

  • past.csv - contains revenue per day with ad spend per day
  • future.csv - contains ad spends, but doesn't contain revenue

Tasks:

  • working source code in Python
  • result dataset in csv format - it should contain date, actual and fitted values for past and future
  • plot of historical vs. fitted revenue + predicted future revenue
  • plot of prophet trends decomposition (see attached example)

Results:

  • For the revenue predicting Prophet model was used with the tuned hyperparameters
  • As a result of hyperparameters tunning root mean square error (rmse) dropped from more then $700 to $30
  • From the forecast could be found that trend has a negative impact throughout time which indicated a need for future modification of the marketing strategy

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published