For example,

I have a category A called fruit, its children are in category B, for example :

  • apple
  • banana
  • ....

The whole category A have been already sold for 3 years , but now we have a new category B fruit call cherry, which only sold for 3 week .

I have thought to transfer trend/seasonal to children , scaled down by the 7 days mean salecount, this method is too simple with many problems . I don't know what is the proper name for this kind of problem, it is different from transfer learning in deep learning area . Is there any smarter way to transfer knowledge in category A to this new category B ? Any thesis are welcome .

  • $\begingroup$ Maybe hierarchical time series? Search this site. $\endgroup$ Jul 24, 2020 at 17:25
  • $\begingroup$ @kjetil b halvorsen I have 70+ catA, 500+ cat B, 7000+ cat C . It would need tremendous calculation resouse to build a hierarchical model , so I just want one to one transfer . $\endgroup$
    – Mithril
    Jul 25, 2020 at 4:24


Your Answer

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