I am using python and Xgboost. I have features: activity and location and time stamps of when the activity occurred.

I want to predict day of week. Is this straight forward, ie y=day of week, X={activity, location}, or am I being naive and I need to do fancy time series things? I'd also like to predict time of day.

  • 1
    $\begingroup$ Do you have a finite list of activities? How many different activities are there? What is location? Is it lattitude and longitude, or you have some finite list of locations (how long), or something else? $\endgroup$ – user31264 Jun 25 '19 at 16:02
  • $\begingroup$ Small number of activities: either around 10 or 200, not sure which I will use. Location is a region: just an arbitrary number denoting a geographical area. $\endgroup$ – schoon Jun 26 '19 at 16:51
  • $\begingroup$ How many regions are there, and what is the size of training set? $\endgroup$ – user31264 Jun 28 '19 at 6:52
  • $\begingroup$ Sorry I missed your comment. Around 100 regions although possibly it could be as many as 1000. Training set is around 150,000. $\endgroup$ – schoon Jul 19 '19 at 14:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.