1
$\begingroup$

I'm new to data science and I'm working on a regression problem. My question is the index of my data which is of type "Date time index" plays a part in regression? I mean is it Okay if i drop the index ?

$\endgroup$
  • $\begingroup$ Please explain more about the problem you are trying to solve (features and target) $\endgroup$ – Alireza Zolanvari Mar 16 '19 at 11:43
  • $\begingroup$ i have hourly data of a smart grid and i want to predict hourly system load $\endgroup$ – imtiaz ul Hassan Mar 16 '19 at 17:17
1
$\begingroup$

You can take the date into account, but you should convert it into numeric value. Suggestions: 1) Consider "datetime" package in python to change it to numeric, or some other techniques such as indexing with date-time, and converting the index value to numeric, float, etc. I don't have your exact case, so cannot say specifically. or 2) If you have time series data, then in many cases ARMAX, ARIMAX, Seasonal ARIMAX models can do a really good job.

$\endgroup$

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.