Let's say I have a set of time series data with 32 time steps.

My goal is to predict what the data value would be for the next time step, given data for 30 previous time steps. Would it be okay to split the dataset with overlapping parts as follows?

train X set : index 1~30  // train Y set : index 31 

test X set : index 2~31 // test Y set : index 32

P.S.: Set aside the dev set for now.

  • 2
    $\begingroup$ Check this answer stats.stackexchange.com/a/346196/89649 $\endgroup$ – Skander H. Aug 11 '18 at 13:56
  • $\begingroup$ @Alex Thanks for the comment. But I was looking for a response that answers the question, 'is it okay to split the train/test with overlapping portions?' $\endgroup$ – HyeongGyu Froilan Choi Aug 11 '18 at 14:05
  • 1
    $\begingroup$ @HyeongGyuFroilan your approach is okay. Follow the links to the Rob Hyndman post in the answer that I indicated. $\endgroup$ – Skander H. Aug 11 '18 at 23:58