Data Science Fellows Projects 2019

Multivariate Time series forecasting

Project by Shalom

Abstract

The project goal was to accurately predict the amount of printing volume per printer per day.

 

Challenges

  • Learning to find the correct time series model to accurately predict the printing volume.
  • Dealing with large quantities of data.

 

Achievements (according to KPIs)

  • Improved the naïve model by 6%.
  • The improvement is by comparing average absolute error printed volume divided by the mean printed volume.
  • The naïve model had a 53% error ratio.
  • My model had a 47% error ratio.

 

Share this post

Share on facebook
Share on twitter
Share on linkedin
Share on email