1
$\begingroup$

Stats model or any other machine learning python packages for doing sequence classification(that can be multi class) and sequence prediction (Both next step and regression).

PS : Input data will be n number of sequences with their respective target (class/regression)

Also each sequences having finite number of events like E1,E2 ..

$\endgroup$
2
  • $\begingroup$ Are the sequences of fixed length? $\endgroup$ Commented May 17, 2019 at 8:26
  • $\begingroup$ nope. It will be of varying length. Can we pad.? $\endgroup$ Commented May 17, 2019 at 9:11

1 Answer 1

1
$\begingroup$

Have you thought of using Conditional Random Fields? I had used it for my thesis some years ago and papers from Prof. Sunita Sarawagi were a great resource for me. You can use the CRFsuite library which also offers a python wrapper.

$\endgroup$
3
  • $\begingroup$ Does this handles if target is regression.? Because i have already tried seqlearn, which uses HMM and classifies it. But the real problem here is conversion of class specific functions to regression value $\endgroup$ Commented May 17, 2019 at 9:17
  • $\begingroup$ Do you really have a regression problem to solve or is it more of a multi-class classification problem? $\endgroup$
    – mari
    Commented May 17, 2019 at 11:44
  • $\begingroup$ yes it involves regression too $\endgroup$ Commented May 17, 2019 at 19:18

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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