I need to test different datasets as well as different algorithm implementations. The current workflow looks like:

  • Perform feature extraction from train set
  • Train classifier on this features
  • Feed this classifier to production code
  • Run production code on a test set, feeding samples one by one
  • Take results of production code and convert to the same format as test dataset
  • Show statistics

How to automate this as much as possible? I currently use a bunch of bash/python scripts. I've looked through sklearn pipelines, but not sure they would give any benefit except for running functions in a line.


1 Answer 1


Maybe it is a bit too much just for this use case, but I have good experience with airflow.

It is an Apache project and quite helpful to automate some stuff.

Otherwise there are a lot of commercial platform helping you. Just google for data science platform.


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.